Clinical definitions of RA remission

Short et al. published a definition of RA remission as early as in 1948, stating that patients were classified as in remission “if the disease was inactive, the patients were asymptomatic and examination of the joints was negative except for residual deformity.” The first formal definition of remission in RA was developed by an ARA (later ACR) subcommittee and published in 1981 . These ‘preliminary’ criteria for clinical remission have remained the most widely applied criteria until recently. In a publication, Pinals et al. stated that “‘complete remission’ implies the total absence of all articular and extra-articular inflammation and immunologic activity related to RA.” The aim of the development of this standard definition of ‘complete clinical remission’ was to dispel the confusion regarding the term ‘remission’ at the time, and the definition was also intended as a potential ‘end’ point for therapeutic trials. To be in remission according to the 1981 ARA criteria, five or more of the following six requirements must be fulfilled for at least 2 consecutive months: Duration of morning stiffness not exceeding 15 min, no fatigue, no joint pain (by history), no joint tenderness or pain on motion, no soft tissue swelling in joints or tendon sheaths and erythrocyte sedimentation rate (ESR) less than 30 mm h ⁻¹ for a female or 20 mm h ⁻¹ for a male. Due to their stringency, the requirement of 2-month duration and inclusion of elements not part of the ACR core set for RA, these criteria have not been widely used in clinical trials or clinical practice. Many modifications have also been developed, for example, omitting one or more of the six requirements or the requirement of 2 consecutive months.

Remission definitions based on composite disease activity indices

Remission definitions based on composite disease activity indices have gained widespread use during the past 10–15 years. A remissions cut-off of <1.6 was proposed for the original DAS based on a study relating DAS values to ARA remission status (a modified version of the ARA criteria without fatigue), and this cut-off has been used since . For DAS28, a remission cut-off of <2.6 was proposed based on a formula calculating DAS28 from the original DAS . This cut-off was later tested against the ARA remission criteria and found to correspond well . Several studies have compared the original DAS and DAS28 remission cut-offs to each other and to the ARA remission criteria . Especially the DAS28 < 2.6 cut-off has been criticised for not being stringent enough, and it has been argued that the reduced 28 joint counts omitting the feet limits the validity of this remission definition . Lowering the cut-off for DAS28 remission to <2.3 or <2.4 has been suggested . Studies of agreement between the original DAS and DAS28 remission cut-offs have yielded variable results . Whereas remission is often defined as total absence of signs and symptoms of disease activity, the DAS cut-offs rather correspond to a state of low disease activity, ‘near-remission’ or ‘partial remission’.

Remission cut-offs have also been developed for SDAI (≤3.3) and CDAI (≤2.8) . These cut-offs are consistently more stringent than the DAS28 cut-off, and due to differences in construction of the indices, the SDAI and CDAI remission cut-offs allow for fewer residual swollen joints . Furthermore, remission cut-offs have been proposed for the solely patient-derived indices Patient Activity Scale (PAS), based on the Health Assessment Questionnaire (HAQ) and visual analogue scales for pain and patient global items and Routine Assessment of Patient Index Data 3 (RAPID), based on function, pain and patient global items from the Multi Dimensional Health Assessment Questionnaire (MDHAQ), both developed for use in clinical practice .

New ACR/EULAR remission criteria

With remission becoming an achievable goal for more patients, the need for consensus on how to define remission is increasingly pertinent. During the last few years ACR and EULAR have collaborated to re-evaluate the concept of remission with the aim to reach consensus on a new, stringent definition of remission . Through an elaborate process, including testing of predictive validity versus radiographic progression and deterioration of physical function, several candidate remission definitions were tested. Two definitions of remission were agreed upon, and it was recommended that every future RA clinical trial should include one of the two remission definitions as an outcome. The new ACR/EULAR remission criteria include: (1) A Boolean-based definition: Tender joint count, swollen joint count, C-reactive protein (CRP, in mg dl ⁻¹ ) and patient global assessment (0–10 scale) all ≤1 and (2) An index-based definition: SDAI ≤3.3 . It was decided that the criteria could be applied when only 28 joint counts are available, but the recommendation from the group was to perform more extensive joint counts including the ankles and forefeet .

Imaging modalities in the assessment of remission

Imaging modalities capture two main features in RA – structural damage and inflammation ( Table 1 ). To define remission by imaging, we first have to clarify which of these two we want to examine. Structural damage can be assessed by conventional radiography (CR), computed tomography (CT), ultrasonography and magnetic resonance imaging (MRI), while inflammation can mainly be assessed by ultrasonography and MRI. In this chapter, we focus on inflammation and imaging remission defined as absence of inflammation – both clinical and sub-clinical. An alternative definition of imaging remission is ‘no progression of structural damage’, an interesting concept that will only be touched upon in the current review.

Imaging of inflammatory activity is gaining influence in rheumatology clinical practice, and it is important that advanced imaging modalities obtain correct information of the degree of inflammatory activity, in addition to detection of early erosive changes. The most promising emerging imaging modalities in this respect are musculoskeletal ultrasonography and MRI, with conventional radiography as the reference method to evaluate structural changes.

Ultrasonography

Ultrasonography has a rapidly growing application in the investigation and management of rheumatic diseases. It has a number of practical advantages over other advanced imaging techniques such as MRI including low cost, good accessibility including the possibility to perform examinations in the outpatient clinic as an integrated part of the patient consultation and the ability to scan multiple joints in a relatively short period of time . Another advantage is the simultaneous imaging of bone and soft tissue, without the need for intravenous contrast. Ultrasonography may be performed routinely by the properly trained rheumatologist in a clinical setting to obtain information about the amount of synovitis (greyscale, GS) as well as the degree of vascularisation, assessed by the use of power Doppler (PD) . Scanning of joints and soft tissues may give information about synovitis, tenosynovitis, PD activity, effusions, erosions and degenerative changes . The validity of ultrasonography for assessment of synovitis and erosive changes has been shown against MRI and CT as gold standards .

More specifically, the PD signal has proved to be a simple tool for short-term monitoring of synovial vascularity changes in RA patients, and it is also a promising examination for prognostification in early RA . Several studies have also indicated that ultrasonography may be an important assessment in patients in apparent clinical remission; this will be reviewed later in this chapter.

Magnetic resonance imaging

MRI is a non-invasive tomographic imaging technique that yields cross-sectional images in any plane, without morphologic distortion or magnification. Projectional superimposition, which is a problem with conventional radiography, is avoided with MRI because of its multiplanar capabilities. MRI allows simultaneous examination of all components of the diarthrodial joint, including soft tissues, articular cartilage and bone, without ionising radiation and adverse effects. MRI is the only imaging technique that can visualise bone marrow oedema (BME, also known as osteitis), one of the strongest predictors of future joint damage . While conventional radiography visualises the structural changes that are the cumulative results of preceding disease activity , MRI allows direct visualisation and assessment of synovitis, the primary lesion in RA. MRI captures sub-clinical inflammation in RA, that is, tenosynovitis, synovitis and osteitis, and results from studies assessing MRI in RA remission will be reviewed later.

Novel imaging modalities

Positron emission tomography (PET) is a nuclear medicine imaging technique that visualises functional processes in the body. In modern PET scanners, CT or MRI scans are used to produce three-dimensional (3D) imaging with anatomic and metabolic information (i.e., what the structure is and what it is doing biochemically). [18F]-fluorodeoxyglucose (FDG) is a tracer for glucose metabolism that can be found in inflammatory tissue, including macrophages, capillaries and fibroblasts. RA inflammation is characterised by fibrovascular proliferation with high FDG uptake, making the method useful in the assessment of RA joints. The technique has also been used in monitoring of treatment effects . The first true hybrid PET-MRI examination of the hand in early RA was reported by Miese et al. Increased 18F-FDG uptake was present in MRI-detected synovitis and tenosynovitis, and the authors concluded that in RA, hybrid 18F-FDG PET-MRI of the hand is technically feasible and can potentially be used to directly visualise inflammation. These early studies are promising, but further research is needed to determine the use in RA, and especially in RA remission.

Indocyanine green (ICG)-enhanced fluorescence optical imaging (FOI) is a novel technology offering sensitive imaging detection of inflammatory changes in subjects with arthritis. In a recent study of patients with arthritis in the hand , FOI was more sensitive than clinical examination and showed good agreement with clinical examination, PD ultrasonography (PDUS) and MRI. FOI seemed to be as informative as 1.5 T MRI and ultrasonography for the detection of synovitis and tenosynovitis, but this needs to be confirmed in further studies, and especially the specificity for detection of synovitis needs to be established.

Importance of inflammatory imaging findings in active disease

The advent of potent, but expensive, biologic therapies has highlighted the need for reliable prognostic markers to better tailor treatment options and strategies to the individual patient. Many factors are known to predict prognosis, including (but not limited to) anti-citrullinated protein, antibodies status, rheumatoid factor status, age, gender, genes, inflammatory markers (ESR/CRP) and erosions . In addition, modern imaging findings have been shown to have predictive value in active disease.

In a study assessing erosive progression in early RA, Bøyesen et al. found ultrasonography GS synovitis and MRI BME to be independent predictors of 1-year MRI erosive progression . In the same cohort, ultrasonography-assessed extensor carpi ulnaris tenosynovitis was an independent predictor of development of erosive joint damage . Naredo et al. showed that time-integrated PDUS demonstrated a strong correlation with radiographic progression in a small study of 42 early RA patients . Ultrasonography quantitative PD scores have been found to predict radiographic progression in individual finger joints in patients with active RA .

Several studies have shown encouraging data regarding the predictive value of MRI findings for future erosive progression of conventional radiographic damage. Some examined the short-term predictive value of MRI findings after 1-year or 2 years ; other studies had 5 , 6 or 10 years’ follow-up. In all studies, MRI findings were found to be significant predictors of progression of erosions. In two small studies of early RA patients, MRI BME at baseline was the strongest individual predictor of MRI erosions at 1 and 2 years’ follow-up . MRI synovitis showed a clear relationship with the development of subsequent MRI erosive damage in another study, but this study did not include conventional radiographs, generally regarded as the gold standard for assessing structural damage . Haavardsholm et al. showed that according to the Outcome Measures in Rheumatology Clinical Trials (OMERACT) RA MRI Scoring system (RAMRIS) the MRI BME score was an independent predictor of structural progression on CR and MRI in an inception cohort of 84 early RA patients . In the randomised controlled Cyclosporine, Methotrexate, Steroid in RA (CIMESTRA) trial of 130 RA patients, baseline RAMRIS MRI BME score of metacarpophalangeal (MCP) and wrist joints was the strongest independent predictor of radiographic progression in hands, wrists and forefeet after 2 years . The 5-year follow-up of these patients confirmed a long-term predictive ability of BME . A systematic review by Suter et al. evaluated the prognostic value of MRI in patients with undifferentiated inflammatory arthritis (UA) or early RA. MRI findings were significantly associated with subsequent radiographic progression in only two of the nine studies in which uniform treatment was used throughout. The authors concluded that data evaluating MRI for the prognosis of early RA are currently inadequate to justify widespread use of the technology for this purposes, although MRI BME may be predictive of progression in certain RA populations. However, the Suter et al. review did not, when reaching this conclusion, take into consideration differences in design, outcomes and quality of the included studies. In a systematic review of the value of MRI and ultrasonography in undifferentiated arthritis, Machado et al concluded that MRI bone marrow edema and combined synovitis and erosion pattern seem useful in predicting development of RA in patients with undifferentiated peripheral inflammatory arthritis .

Thus, several studies show that MRI BME should be recognised as an important prognostic factor in RA, at least in active disease, and that MR synovitis may be of prognostic value. Studies assessing the predictive value of ultrasonography findings in active disease are sparse, and there is a general lack of studies evaluating the independent contribution of ultrasonography findings, but the available studies report promising findings. Both ultrasonography and MRI findings in patients with low disease activity or in remission show promising results, and the main studies addressing this are reviewed below.

Clinical definitions of RA remission

Short et al. published a definition of RA remission as early as in 1948, stating that patients were classified as in remission “if the disease was inactive, the patients were asymptomatic and examination of the joints was negative except for residual deformity.” The first formal definition of remission in RA was developed by an ARA (later ACR) subcommittee and published in 1981 . These ‘preliminary’ criteria for clinical remission have remained the most widely applied criteria until recently. In a publication, Pinals et al. stated that “‘complete remission’ implies the total absence of all articular and extra-articular inflammation and immunologic activity related to RA.” The aim of the development of this standard definition of ‘complete clinical remission’ was to dispel the confusion regarding the term ‘remission’ at the time, and the definition was also intended as a potential ‘end’ point for therapeutic trials. To be in remission according to the 1981 ARA criteria, five or more of the following six requirements must be fulfilled for at least 2 consecutive months: Duration of morning stiffness not exceeding 15 min, no fatigue, no joint pain (by history), no joint tenderness or pain on motion, no soft tissue swelling in joints or tendon sheaths and erythrocyte sedimentation rate (ESR) less than 30 mm h ⁻¹ for a female or 20 mm h ⁻¹ for a male. Due to their stringency, the requirement of 2-month duration and inclusion of elements not part of the ACR core set for RA, these criteria have not been widely used in clinical trials or clinical practice. Many modifications have also been developed, for example, omitting one or more of the six requirements or the requirement of 2 consecutive months.

Remission definitions based on composite disease activity indices

Remission definitions based on composite disease activity indices have gained widespread use during the past 10–15 years. A remissions cut-off of <1.6 was proposed for the original DAS based on a study relating DAS values to ARA remission status (a modified version of the ARA criteria without fatigue), and this cut-off has been used since . For DAS28, a remission cut-off of <2.6 was proposed based on a formula calculating DAS28 from the original DAS . This cut-off was later tested against the ARA remission criteria and found to correspond well . Several studies have compared the original DAS and DAS28 remission cut-offs to each other and to the ARA remission criteria . Especially the DAS28 < 2.6 cut-off has been criticised for not being stringent enough, and it has been argued that the reduced 28 joint counts omitting the feet limits the validity of this remission definition . Lowering the cut-off for DAS28 remission to <2.3 or <2.4 has been suggested . Studies of agreement between the original DAS and DAS28 remission cut-offs have yielded variable results . Whereas remission is often defined as total absence of signs and symptoms of disease activity, the DAS cut-offs rather correspond to a state of low disease activity, ‘near-remission’ or ‘partial remission’.

Remission cut-offs have also been developed for SDAI (≤3.3) and CDAI (≤2.8) . These cut-offs are consistently more stringent than the DAS28 cut-off, and due to differences in construction of the indices, the SDAI and CDAI remission cut-offs allow for fewer residual swollen joints . Furthermore, remission cut-offs have been proposed for the solely patient-derived indices Patient Activity Scale (PAS), based on the Health Assessment Questionnaire (HAQ) and visual analogue scales for pain and patient global items and Routine Assessment of Patient Index Data 3 (RAPID), based on function, pain and patient global items from the Multi Dimensional Health Assessment Questionnaire (MDHAQ), both developed for use in clinical practice .

New ACR/EULAR remission criteria

With remission becoming an achievable goal for more patients, the need for consensus on how to define remission is increasingly pertinent. During the last few years ACR and EULAR have collaborated to re-evaluate the concept of remission with the aim to reach consensus on a new, stringent definition of remission . Through an elaborate process, including testing of predictive validity versus radiographic progression and deterioration of physical function, several candidate remission definitions were tested. Two definitions of remission were agreed upon, and it was recommended that every future RA clinical trial should include one of the two remission definitions as an outcome. The new ACR/EULAR remission criteria include: (1) A Boolean-based definition: Tender joint count, swollen joint count, C-reactive protein (CRP, in mg dl ⁻¹ ) and patient global assessment (0–10 scale) all ≤1 and (2) An index-based definition: SDAI ≤3.3 . It was decided that the criteria could be applied when only 28 joint counts are available, but the recommendation from the group was to perform more extensive joint counts including the ankles and forefeet .

Imaging modalities in the assessment of remission

Imaging modalities capture two main features in RA – structural damage and inflammation ( Table 1 ). To define remission by imaging, we first have to clarify which of these two we want to examine. Structural damage can be assessed by conventional radiography (CR), computed tomography (CT), ultrasonography and magnetic resonance imaging (MRI), while inflammation can mainly be assessed by ultrasonography and MRI. In this chapter, we focus on inflammation and imaging remission defined as absence of inflammation – both clinical and sub-clinical. An alternative definition of imaging remission is ‘no progression of structural damage’, an interesting concept that will only be touched upon in the current review.

Imaging of inflammatory activity is gaining influence in rheumatology clinical practice, and it is important that advanced imaging modalities obtain correct information of the degree of inflammatory activity, in addition to detection of early erosive changes. The most promising emerging imaging modalities in this respect are musculoskeletal ultrasonography and MRI, with conventional radiography as the reference method to evaluate structural changes.

Ultrasonography

Ultrasonography has a rapidly growing application in the investigation and management of rheumatic diseases. It has a number of practical advantages over other advanced imaging techniques such as MRI including low cost, good accessibility including the possibility to perform examinations in the outpatient clinic as an integrated part of the patient consultation and the ability to scan multiple joints in a relatively short period of time . Another advantage is the simultaneous imaging of bone and soft tissue, without the need for intravenous contrast. Ultrasonography may be performed routinely by the properly trained rheumatologist in a clinical setting to obtain information about the amount of synovitis (greyscale, GS) as well as the degree of vascularisation, assessed by the use of power Doppler (PD) . Scanning of joints and soft tissues may give information about synovitis, tenosynovitis, PD activity, effusions, erosions and degenerative changes . The validity of ultrasonography for assessment of synovitis and erosive changes has been shown against MRI and CT as gold standards .

More specifically, the PD signal has proved to be a simple tool for short-term monitoring of synovial vascularity changes in RA patients, and it is also a promising examination for prognostification in early RA . Several studies have also indicated that ultrasonography may be an important assessment in patients in apparent clinical remission; this will be reviewed later in this chapter.

Magnetic resonance imaging

MRI is a non-invasive tomographic imaging technique that yields cross-sectional images in any plane, without morphologic distortion or magnification. Projectional superimposition, which is a problem with conventional radiography, is avoided with MRI because of its multiplanar capabilities. MRI allows simultaneous examination of all components of the diarthrodial joint, including soft tissues, articular cartilage and bone, without ionising radiation and adverse effects. MRI is the only imaging technique that can visualise bone marrow oedema (BME, also known as osteitis), one of the strongest predictors of future joint damage . While conventional radiography visualises the structural changes that are the cumulative results of preceding disease activity , MRI allows direct visualisation and assessment of synovitis, the primary lesion in RA. MRI captures sub-clinical inflammation in RA, that is, tenosynovitis, synovitis and osteitis, and results from studies assessing MRI in RA remission will be reviewed later.

Novel imaging modalities

Positron emission tomography (PET) is a nuclear medicine imaging technique that visualises functional processes in the body. In modern PET scanners, CT or MRI scans are used to produce three-dimensional (3D) imaging with anatomic and metabolic information (i.e., what the structure is and what it is doing biochemically). [18F]-fluorodeoxyglucose (FDG) is a tracer for glucose metabolism that can be found in inflammatory tissue, including macrophages, capillaries and fibroblasts. RA inflammation is characterised by fibrovascular proliferation with high FDG uptake, making the method useful in the assessment of RA joints. The technique has also been used in monitoring of treatment effects . The first true hybrid PET-MRI examination of the hand in early RA was reported by Miese et al. Increased 18F-FDG uptake was present in MRI-detected synovitis and tenosynovitis, and the authors concluded that in RA, hybrid 18F-FDG PET-MRI of the hand is technically feasible and can potentially be used to directly visualise inflammation. These early studies are promising, but further research is needed to determine the use in RA, and especially in RA remission.

Indocyanine green (ICG)-enhanced fluorescence optical imaging (FOI) is a novel technology offering sensitive imaging detection of inflammatory changes in subjects with arthritis. In a recent study of patients with arthritis in the hand , FOI was more sensitive than clinical examination and showed good agreement with clinical examination, PD ultrasonography (PDUS) and MRI. FOI seemed to be as informative as 1.5 T MRI and ultrasonography for the detection of synovitis and tenosynovitis, but this needs to be confirmed in further studies, and especially the specificity for detection of synovitis needs to be established.

Importance of inflammatory imaging findings in active disease

The advent of potent, but expensive, biologic therapies has highlighted the need for reliable prognostic markers to better tailor treatment options and strategies to the individual patient. Many factors are known to predict prognosis, including (but not limited to) anti-citrullinated protein, antibodies status, rheumatoid factor status, age, gender, genes, inflammatory markers (ESR/CRP) and erosions . In addition, modern imaging findings have been shown to have predictive value in active disease.

In a study assessing erosive progression in early RA, Bøyesen et al. found ultrasonography GS synovitis and MRI BME to be independent predictors of 1-year MRI erosive progression . In the same cohort, ultrasonography-assessed extensor carpi ulnaris tenosynovitis was an independent predictor of development of erosive joint damage . Naredo et al. showed that time-integrated PDUS demonstrated a strong correlation with radiographic progression in a small study of 42 early RA patients . Ultrasonography quantitative PD scores have been found to predict radiographic progression in individual finger joints in patients with active RA .

Several studies have shown encouraging data regarding the predictive value of MRI findings for future erosive progression of conventional radiographic damage. Some examined the short-term predictive value of MRI findings after 1-year or 2 years ; other studies had 5 , 6 or 10 years’ follow-up. In all studies, MRI findings were found to be significant predictors of progression of erosions. In two small studies of early RA patients, MRI BME at baseline was the strongest individual predictor of MRI erosions at 1 and 2 years’ follow-up . MRI synovitis showed a clear relationship with the development of subsequent MRI erosive damage in another study, but this study did not include conventional radiographs, generally regarded as the gold standard for assessing structural damage . Haavardsholm et al. showed that according to the Outcome Measures in Rheumatology Clinical Trials (OMERACT) RA MRI Scoring system (RAMRIS) the MRI BME score was an independent predictor of structural progression on CR and MRI in an inception cohort of 84 early RA patients . In the randomised controlled Cyclosporine, Methotrexate, Steroid in RA (CIMESTRA) trial of 130 RA patients, baseline RAMRIS MRI BME score of metacarpophalangeal (MCP) and wrist joints was the strongest independent predictor of radiographic progression in hands, wrists and forefeet after 2 years . The 5-year follow-up of these patients confirmed a long-term predictive ability of BME . A systematic review by Suter et al. evaluated the prognostic value of MRI in patients with undifferentiated inflammatory arthritis (UA) or early RA. MRI findings were significantly associated with subsequent radiographic progression in only two of the nine studies in which uniform treatment was used throughout. The authors concluded that data evaluating MRI for the prognosis of early RA are currently inadequate to justify widespread use of the technology for this purposes, although MRI BME may be predictive of progression in certain RA populations. However, the Suter et al. review did not, when reaching this conclusion, take into consideration differences in design, outcomes and quality of the included studies. In a systematic review of the value of MRI and ultrasonography in undifferentiated arthritis, Machado et al concluded that MRI bone marrow edema and combined synovitis and erosion pattern seem useful in predicting development of RA in patients with undifferentiated peripheral inflammatory arthritis .

Thus, several studies show that MRI BME should be recognised as an important prognostic factor in RA, at least in active disease, and that MR synovitis may be of prognostic value. Studies assessing the predictive value of ultrasonography findings in active disease are sparse, and there is a general lack of studies evaluating the independent contribution of ultrasonography findings, but the available studies report promising findings. Both ultrasonography and MRI findings in patients with low disease activity or in remission show promising results, and the main studies addressing this are reviewed below.

Progression of joint damage in clinical remission

Radiographic joint damage is one of the main outcomes in RA and is associated with impaired physical function . Joint damage is less common in RA remission, but still of importance – if an RA patient achieves clinical remission but has ongoing significant joint damage, the situation will not be satisfactory to the clinician. This is also illustrated by the development of the 2011 ACR/EULAR remission criteria, where candidate criteria were tested for their ability to predict a good future radiographic and functional outcome .

A summary of studies of joint damage in RA remission identified no observational studies or randomised controlled trials (RCTs) in which patients in remission did not show any radiographic progression . In the four observational studies identified continued radiographic progression despite clinical remission was found in 19–54% of patients over 3–5 years . Three studies assessed erosive progression in patients achieving remission compared to other cases: one study reported lower erosive progression with sustained remission (19% vs. 72%) , whereas two other studies found no differences. In the RCTs included in this review, patients showed less progression of joint damage in remission when treated with combination therapies compared to monotherapies .

In the development of the 2011 ACR/EULAR remission criteria, candidate criteria were tested in data from several RCTs. Of patients fulfilling the final Boolean criteria, 23% experienced continued joint damage . A study comparing several remission criteria, including the 2011 ACR/EULAR criteria, in established RA found similar results .

Modern imaging findings in clinical remission

This section reviews the most important studies assessing MRI- and ultrasonography-detected inflammation in RA patients in clinical remission.

Brown et al. examined 107 RA patients receiving disease-modifying antirheumatic drug (DMARD) therapy who were judged by their rheumatologist to be in remission . Clinical, laboratory, functional and quality-of-life assessments were obtained, and hands and wrists were examined with MRI and ultrasonography. Various clinical criteria were applied to determine remission, with 55% of patients fulfilling the 1981 ARA remission criteria and 57% the DAS28 remission criteria. Of the RA patients who were in remission according to the 1981 ARA remission criteria, 96% had synovitis detected by MRI and 81% had synovial hypertrophy on US. Of patients who were in remission according to the DAS28 criteria, 96% had synovitis on MRI and 84% had synovial hypertrophy on US. The number of patients with synovial pathology detectable by MRI or ultrasonography was not significantly different if DAS28 or ARA remission criteria were applied. The authors postulated that this sub-clinical inflammation may explain the observed discrepancy between disease activity and outcome in RA, and concluded that imaging assessment may be necessary for the accurate evaluation of disease status and, in particular, for the definition of true remission. Further, they concluded that the term ‘true remission’ should be reserved for patients who are not only in a state of clinical remission, but also show an absence of synovitis on imaging studies .

In 2008, Brown et al. published a longitudinal evaluation of the clinical remission cohort described above, aiming to determine the structural significance of the sub-clinical inflammation and to investigate whether imaging-detected synovitis could be used to predict subsequent progression of joint damage . All patients underwent complete clinical, laboratory, radiographic, functional, quality-of-life and imaging assessments at baseline and 12 months. Radiographs at both time points were available for 90 patients, and these patients formed the core data set for the analyses. At baseline, 68% of patients had ultrasonography erosions, 89% had evidence of ultrasonography synovial hypertrophy and 63% had increased PDUS signal. The results of the MRI examinations showed that 96% of patients had erosions, 92% had synovitis and 53% had BME. Seventeen of 90 patients experienced deterioration in their total radiographic joint damage scores during the 12-month follow-up period. Only 5% of patients were subject to change in therapy due to increased disease activity throughout this period. At the patient level, the total baseline PD score was found to be a predictor of subsequent progression of joint damage in univariate analyses. For individual joints, ultrasonography synovial hypertrophy, PD and MRI synovitis at baseline were significantly associated with progressive radiographic damage. Ultrasonography PD score in asymptomatic MCP joints at baseline was associated with structural progression over 12 months, and a 12 times higher odds of deterioration in joints with increased PD signal. The authors concluded that sub-clinical joint inflammation detected by modern imaging techniques explains the structural deterioration in RA patients in clinical remission, and that their findings reinforce the utility of imaging for the accurate evaluation of disease status and the prediction of structural outcome .

Scire et al. studied an inception cohort of 106 early RA patients receiving synthetic DMARDs according to a DAS-steered therapeutic protocol over a 24-month period . The evaluation included clinical and laboratory assessments, as well as a systematic ultrasonography assessment of 44 joints. Remission was defined as DAS <1.6 at two consecutive visits 3 months apart. Ninety-five percent of patients in clinical remission according to this definition showed residual GS synovitis, and 41% showed a positive PD signal. A positive PD signal, even in a single joint, proved to be the main predictor of relapse within 6 months. An important finding in this study was the high negative predictive value of the PD assessment. The absence of a PD signal was associated with stable remission in >90% of the cases during the 24-month follow-up, supporting the concept that a more complete remission might be defined by a negative PD finding. The authors suggested that once stable clinical remission has been achieved, a step-down therapeutic strategy should be limited to patients with a negative PD signal, and concluded that ultrasonography PD monitoring in daily practice might lead to increased remission rates and reduce disease relapses and structural damage progression of early RA patients .

Balsa et al. evaluated a cohort of 97 patients with RA to examine the relationship between clinical remission and imaging remission . As in the Brown studies, patients were classified as being in remission based on their treating rheumatologist’s clinical judgement. Patients underwent a complete clinical and laboratory assessment of disease activity using standard methods, as well as imaging by ultrasonography and conventional radiography. Seventy-four patients (76%) were in remission according to the DAS28 cut-off of <2.6, and 43 (44%) patients were in remission according to an SDAI cut-off of ≤3.3. Seventy-three patients (75%) fulfilled a modified version of the ARA criteria for remission. A total of 42 joints were assessed by ultrasonography, and 92 out of the 97 RA patients (95%) displayed synovial hypertrophy while 41 (42%) had a PD signal in at least one joint. The authors defined ‘true remission’ as absence of joints with PD signal, and compared the performance of DAS28 remission and SDAI remission with regard to this outcome. No difference was found in the prevalence of DAS28 remission between patients in ultrasonography remission and those with active disease as defined by US. However, a clear relationship was found for SDAI remission. Balsa et al. concluded that SDAI remission is closer to the concept of an absence of inflammatory activity, but that further studies are required to assess whether these findings are also associated with improved clinical outcomes .

Peluso et al. examined two cohorts of patients in DAS remission to assess the prevalence of ultrasonography remission, possible predictors of ultrasonography remission and how many patients in the two cohorts with and without ultrasonography remission relapsed after 1-year of follow-up . The first cohort included patients with early RA (ERA), while the other cohort consisted of patients with longstanding RA (LSRA). A total of 48 patients with ERA and 46 with LSRA were included; 57% of patients with ERA and 50% of those with LSRA fulfilled the 1981 ARA remission criteria . Ultrasonography examination showed no synovitis in 44% of patients with ERA and in 17% of those with LSRA, and early disease was associated with clinical ultrasonography remission. Twenty percent of patients with RA who had a negative PD signal at the baseline ultrasonography evaluation had a flare during the 12-month follow-up period compared with 47% of patients who had a positive PD signal. The authors suggest that a combined approach of both clinical and ultrasonography evaluation should be used to define true remission in patients with RA .

In a study by Saleem et al., 128 RA patients with sustained DAS28 ≤ 2.6 for at least 6 months were classified according to clinical-(DAS28 and SDAI) and ultrasonography-based remission definitions . The patients were treated with either synthetic DMARDs alone ( n = 66) or in combination with a tumour necrosis factor blocker ( n = 62). Of the 640 examined joints, 5% had moderate or severe PD activity (defined as a PD score >2), 8% were clinically swollen and 1% tender. Moderate or severe PD activity was present in 21% of patients fulfilling DAS28 remission criteria, 15% fulfilling ARA criteria and 19% fulfilling SDAI criteria. The application of more stringent remission cut-offs resulted in reduced signs and symptoms of inflammation, but the percentage of joints with PD activity was not reduced, even in those without signs or symptoms. These data suggest that clinical criteria are insensitive to detect low but clinically relevant levels of inflammation, and the authors concluded that their study provided additional evidence to support the use of imaging to ensure accurate disease activity assessment in patients in a low disease activity state or remission .

Døhn et al. followed 52 biologically naive RA patients with erosive disease with repeated imaging (MRI, ultrasonography and radiography), clinical and biochemical assessments during a combination therapy with adalimumab and methotrexate . Examinations took place at the initiation of combination therapy and after 6 and 12 months, with 38% of patients reaching DAS28 remission at 12 months. All patients had MRI-detectable synovitis and GS synovitis at 12 months, while 63% of patients had BME and 87% PD synovitis. After 12 months, MRI synovitis was detected in 95% of joints and MRI bone oedema in 20% of bones, whereas on ultrasonography, GS and PD synovitis were found in 45% and 26% of the joints examined, respectively. Only a minority of the patients was in clinical remission, but the percentage of patients with positive imaging findings of inflammation was high. No overall progression of structural damage assessed by MRI and ultrasonography was seen over that same period, but patients with definite erosive progression at 12 months on CT (an increase in erosive joint damage > smallest detectable change) had significantly higher baseline MRI bone oedema scores than non-progressors.

In a study by Gandjbakhch et al., 85 RA patients in remission or low disease activity were followed up for one year to identify predictive factors of structural progression assessed by MRI . Structural progression was defined as a change in the RAMRIS erosion score between baseline and one year greater than the smallest detectable difference. At baseline, a majority of patients displayed inflammatory activity on MRI: 87% had at least one joint with RAMRIS synovitis and 23% at least one location of BME. BME at baseline was predictive of structural progression on MRI . This finding confirms the prognostic role of BME with regard to subsequent bone erosions, even in RA patients in remission or low disease activity.

In a multicentre study coordinated by the OMERACT MRI in the inflammatory arthritis group, data from six cohorts were collected . RA patients in clinical remission ( n = 213) or low disease activity ( n = 81) with available MRI and conventional radiograph scores were included. All MR images were assessed according to the RAMRIS scoring system. MRI inflammatory activity in the wrist and/or MCP joints was observed in the majority of patients. Synovitis and BME were more commonly observed in the wrist (90% synovitis and 31% BME) than in the MCP joints (81% synovitis and 16% BME). No difference was observed between MRI characteristics of patients in clinical remission or LDA, regardless of definition (DAS28, CDAI or SDAI) applied. However, a trend towards lower frequencies of BME was observed when SDAI and CDAI remission criteria were used compared to DAS28 criteria. The authors conclude that sub-clinical inflammation was identified by MRI in the majority of RA patients in clinical remission or a low disease activity state . This project is still ongoing, and preliminary follow-up data were presented at the EULAR congress in Berlin 2012, showing that a high RAMRIS synovitis score was a strong predictor of radiographic progression in rheumatoid factor-positive RA patients in clinical remission and low disease activity .

Saleem et al. evaluated the occurrence of flare in clinical remission, and studied baseline factors associated with disease flare over a 1-year period . Flare was defined as any increase in disease activity requiring a change in therapy. At baseline, 52 patients (56%) fulfilled DAS28 remission criteria, 31 patients (33%) fulfilled SDAI remission criteria, 50 patients (54%) fulfilled the modified 1981 ARA remission criteria and 13 patients (14%) fulfilled 2011 ACR/EULAR remission criteria. Ultrasonography of the dominant hand and wrist showed GS synovial hypertrophy in 89% of patients and PD activity in 62%. During the follow-up period, 26% of patients (24/93) in remission experienced a flare. Fulfilment of remission criteria was not associated with a reduced likelihood of flare, but increased baseline ultrasound PD activity and functional disability were independently associated with risk of flare. Patients who flared had significantly worse clinical and functional outcomes at 12 months compared with patients in sustained remission. The presence of PD activity was the most accurate determinant of flare in RA patients in remission, and in patients with ultrasound-detected PD activity at baseline the odds of a flare of disease were approximately four times as high as the odds for patients with no PD activity. The authors suggest that based on these results, ultrasound could form an important part of remission assessment in RA, and argue that imaging criteria (absence of PD signal) may be superior to fulfilment of clinical remission criteria with regard to prediction of sustained remission .

In another study aiming to assess flare in RA remission, published by Foltz et al. , the objective was to compare the ability of ultrasonography and MRI to predict relapse and radiographic progression. A total of 85 RA patients in either remission or low disease activity were included in the study. During the 1-year follow-up, patients underwent clinical assessments every 3 months and radiographic assessments at baseline and 12 months. Ultrasonography and MRI was performed at baseline. Relapse was defined as a DAS of >2.4. Baseline ultrasonography findings revealed 74 patients (87%) with GS-positive synovitis and 24 (28%) with PD-positive synovitis, while the MRI examinations showed that 82 patients (97%) had synovitis and 27 patients (32%) had BME. One year later, 26 of the 85 patients (31%) had disease relapse, and in 9 of the 85 patients (11%) progression of radiographic joint damage had occurred. The baseline number of joints with PD synovitis predicted relapse (adjusted odds ratio (OR) 6.3), and the baseline PD synovitis grade predicted radiographic joint damage progression (adjusted OR 1.4). Low-field MRI was not predictive of the outcomes. The authors concluded that PD signals on ultrasonography can predict relapse or radiographic progression in RA patients in remission or low disease activity, and that the clinical implications of these findings could be that adjustment of therapy and close monitoring are necessary in these patients, whereas tapering of treatment might not be recommendable .

Yoshimi et al. assessed 31 RA patients in DAS28 remission for at least 2 months . At baseline, PD activity was observed in 20 patients (65%). Twenty-two RA patients maintained clinical remission during the 2-year follow-up period. Radiographic progression was detected in seven (32%) patients in persistent clinical remission. This progression was strongly associated with the total PD score at baseline, with all patients showing radiographic progression having a total PD score of ≥2 at baseline, while none of the patients with a total PD score of ≤1 displayed any radiographic progression. They conclude that these results strongly suggest that ultrasonography-assessed imaging remission may be essential to prevent progressive joint destruction .

Is inflammation detected by modern imaging techniques in RA remission important?

From the studies summarised above, it is clear that sub-clinical inflammation is present in a significant proportion of RA patients in clinical remission, regardless of the definition of remission. Even when applying the strict 1981 ARA remission criteria, almost all patients had inflammatory activity detectable by imaging . The longitudinal studies have suggested that sub-clinical inflammation detected by MRI and ultrasonography may be a prognostic marker of further radiographic progression or flare of disease activity . It should be noted that several of these studies were performed in the same two research groups, and it is possible that some patients were included in more than one cohort. The common conclusion of these studies is that modern imaging adds significantly to clinical assessments in this patient group, and might even provide more important prognostic information in RA remission than composite disease activity measures.

With the growing number of RA patients reaching remission, often in early stages of the disease, the need for improved knowledge on how to best treat these patients is increasing. Tapering of DMARD treatment after prolonged periods of sustained remission is desirable, and these studies indicate that modern imaging can provide important additional information that might help inform this decision .