Abstract
Co-morbid conditions are common in patients with rheumatoid arthritis (RA). Although the presence of co-morbid conditions can be assessed using standardised indexes such as the Charlson index, most clinicians prefer to simply record their presence. Some co-morbidities are causally associated with RA and many others are related to its treatment. Irrespective of their underlying pathogenesis, co-morbidities increase disability and shorten life expectancy, thereby increasing both the impact and mortality of RA. Cardiac co-morbidities are the most crucial, because of their frequency and their negative impacts on health. Treatment of cardiac risk factors and reducing RA inflammation are both critical in reducing cardiac co-morbidities. Gastrointestinal and chest co-morbidities are both also common. They are often associated with drug treatment, including non-steroidal anti-inflammatory drug and disease-modifying drugs. Osteoporosis and its associated fracture risk are equally important and are often linked to long-term glucocorticoid treatment. The range of co-morbidities associated with RA is increasing with the recognition of new problems such as periodontal disease. Optimal medical care for RA should include an assessment of associated co-morbidities and their appropriate management. This includes risk factor modification where possible. This approach is essential to improve quality of life and reduce RA mortality. An area of genuine concern is the impact of treatment on co-morbidities. A substantial proportion is iatrogenic. As immunosuppression with conventional disease-modifying drugs and biologics has many associated risks, ranging from liver disease to chest and other infections, it is essential to balance the risks of co-morbidities against the anticipated benefits of treatment.
In common with other long-term inflammatory diseases, rheumatoid arthritis (RA) has many complex effects on patients. Joint inflammation and damage are the dominant features of RA. However, it also has effects on many other systems. A minority of these are highly specific for RA, the best example being the presence of rheumatoid nodules. However, most are non-specific medical problems. They are broadly collected together as co-morbidities.
At its simplest level, co-morbidity means the existence of two or more diseases in the same person. Although this is a simple concept, there are complex problems in interpreting the importance and associations of co-morbidities. As virtually no one is without any medical problems at all, there is a risk of concluding all patients have co-morbid diseases, which is not helpful. Consequently, some experts have tried to assess the importance of co-morbid diseases by creating an index to define their severity. One example is the Charlson index . Other experts have developed a co-morbidity index focussed on predicting physical function . There are a number of other indices which also assess co-morbidity . None of these approaches is perfect but they all enable some assessment to be made on the overall impact of co-morbidities on RA.
Some co-morbidities are causally associated with RA. Their frequencies and impacts are increased in RA patients. Others co-morbidities, though seen in RA patients, do not have a causal association with RA and their frequencies are not increased. Cardiovascular diseases are an example of the first type of co-morbidity. Many forms of cancer are examples of the second type of co-morbidities. However, irrespective of any causal relationship, all co-morbidities have important roles in worsening the impact of RA.
Co-morbidities and mortality
Patients with RA continue to die prematurely, usually from one or more co-morbid diseases. Gabriel has attributed this excess mortality to three causes . First, RA patients have more serious co-morbid conditions and experience worse outcomes from them, including the risk of premature death. Second, RA patients may not receive optimal preventive care for their co-morbid conditions, which may be overlooked if clinicians focus only on treating RA itself. Finally, systemic inflammation in RA may promote co-morbidity. In essence, the excessive mortality of RA is mainly due to its co-morbidities rather than any direct impact of RA.
When Naz and Symmons reviewed the mortality in RA, they highlighted the marked variation between studies in reporting and finding increased deaths. Some studies show very little increase in mortality. While other studies show marked increases, which can be up to threefold compared to controls. Cardiovascular disease, respiratory disease and infection all contribute to the overall increase in mortality. Although the health of the population as a whole is increasing in Western industrialised societies, a recent report of long-term mortality in an RA cohort initiated in the 1980s showed there is a continuing trend of high mortality in RA, with mortality remaining high for cardiovascular disease and infections .
These studies of the mortality of RA highlight the crucial importance of co-morbidities in determining RA outcomes, and in particular premature death. They also show that, in most patients, RA is an indirect factor in increasing mortality.
Co-morbidities and disability
Those RA patients who have one or more co-morbidities have, on average, more disability than patients without any co-morbidities. This association has been known for many years. It was identified over 20 years ago for work disability . It was subsequently shown in more detail for depression .
Several groups have reported that co-morbidity in general is associated with increased disability in RA. This has been investigated in detail using the health assessment score (HAQ) to measure disability by both Michaud and Wolfe and by Radner et al. . The combined data from these two reports are shown in Fig. 1 . Having one co-morbidity increases the mean HAQ score by approximately 0.2. Multiple co-morbidities increase mean HAQ by approximately 0.8.
There is a complex issue in resolving the nature of this relationship. On the one hand, it could be argued that high HAQ scores indicate severe inflammatory RA, and the systemic inflammation in these patients is likely to drive other diseases, such as cardiovascular disease. Alternatively, patients with multiple co-morbid problems may have several factors influencing their functional state and both their RA and co-morbidities may be contributing to their high disability level. Probably both these explanations are involved in this complex interaction.
Co-morbidities and disability
Those RA patients who have one or more co-morbidities have, on average, more disability than patients without any co-morbidities. This association has been known for many years. It was identified over 20 years ago for work disability . It was subsequently shown in more detail for depression .
Several groups have reported that co-morbidity in general is associated with increased disability in RA. This has been investigated in detail using the health assessment score (HAQ) to measure disability by both Michaud and Wolfe and by Radner et al. . The combined data from these two reports are shown in Fig. 1 . Having one co-morbidity increases the mean HAQ score by approximately 0.2. Multiple co-morbidities increase mean HAQ by approximately 0.8.
There is a complex issue in resolving the nature of this relationship. On the one hand, it could be argued that high HAQ scores indicate severe inflammatory RA, and the systemic inflammation in these patients is likely to drive other diseases, such as cardiovascular disease. Alternatively, patients with multiple co-morbid problems may have several factors influencing their functional state and both their RA and co-morbidities may be contributing to their high disability level. Probably both these explanations are involved in this complex interaction.
Cardiovascular disease
Background
Cardiovascular diseases are potentially the most important co-morbidities in RA. Understanding the cardiovascular risks in RA involves considering three inter-related themes. First, there is the evidence that RA is associated with an increased range of different cardiovascular diseases. This evidence exists for a range of cardiovascular disorders, which have different strengths of evidence for their associated risks. Second, there is the question of whether RA has a specific effect on RA or if its impact on cardiovascular disease is modulated by a change in known risk factors for cardiovascular disease. Finally, there are issues about treatment. A key concern is the possibility that treatments used for RA might increase the risk of associated cardiovascular diseases. An associated concern is whether or not patients with RA have optimal treatments of their cardiovascular risk factors.
Cardiovascular mortality
Evidence that RA patients often die of cardiovascular disease goes back over 50 years to simple observation studies such as that of Reah . The most comprehensive evidence comes from the systematic review of observational studies by Aviña-Zubieta et al. This systemic review evaluated 24 studies that enrolled 111,758 patients who had 22,927 cardiovascular events. Overall, they found a 50% increased risk of cardiovascular deaths in patients with RA. The mortality risks for ischaemic heart disease were increased by 59%. However, there was considerable heterogeneity between studies. Inception cohort studies involving 2175 RA patients did not show an increased cardiovascular mortality. The key findings in this meta-analysis are shown in Fig. 2 , which expresses risks as odds ratios with 95% confidence interval (CI). Of particular interest is the finding that there is no evidence that the increased risk is currently declining. Another systematic review by Meune et al. drew similar conclusions about the increased cardiovascular mortality of RA.
Ischaemic heart disease
There is extensive evidence that patients with established RA have an increased risk of ischaemic heart disease . Another systematic review by Meune et al. identified 10 studies which reported standardised mortality ratios for fatal myocardial infarctions. These enrolled 85,323 patients. The standardised mortality ratios varied from 0.99 to 3.82. Seven studies reported significant increases in fatal myocardial infarctions. Pooled analysis of all the available data showed a standardised mortality ratio of 1.77 (95% CI 1.65, 1.89). The risks of myocardial infarctions were also assessed. The incidence rate ratio for myocardial infarction was evaluated in five studies which enrolled 34,705 patients. The pooled estimate of the incidence risk ratio was 2.10 (95% CI 1.52, 2.89).
There is less evidence about myocardial infarctions in early RA, and particularly in inception cohorts of patients with RA. Holmqvist et al. recently reported the results of a Swedish inception cohort of 7469 RA patients diagnosed between 1995 and 2006. They were compared with a matched general population comparator of 37,024 individuals. Patients and controls were linked to national registers of morbidity and mortality from ischaemic heart disease. During follow-up, 233 patients with RA and 701 controls developed a first myocardial infarction. This gave an overall relative risk of 1.6 (95% CI 1.4, 1.9). The increased risks of myocardial infarction were present within the first 4 years of diagnosis. Interestingly, although the risk of developing ischaemic heart disease and myocardial infarctions was increased, there was no evidence for an increased mortality from myocardial infarction. These findings are shown in Fig. 3 .
There is evidence that RA patients with acute myocardial disease may have a different clinical course. Douglas et al. studied 40 RA patients and the acute with RA and the acute coronary syndrome. Compared to case-matched controls the RA patients had more recurrent cardiac events than controls (58% vs. 30%) and more deaths (40% vs. 15%). All controls presented with chest pain, but this was only seen in 82% of RA patients.
Heart failure
Patients with RA are also more likely to develop heart failure when compared to the population as a whole. A North American study by Nicola et al. evaluated 575 patients with RA and 583 controls. The incidence rates of congestive heart failure were 1.99 and 1.16 cases per 100 person-years in RA patients and controls. After 30 years of follow-up, 34% of RA patients had developed congestive heart failure compared with 25% of controls. The hazard ratio (HR) for congestive heart failure, adjusted for demographics, ischaemic heart disease and cardiovascular risk factors, was 1.87 (95% CI 1.47, 2.39). The risk was higher in rheumatoid factor-positive patients with an HR of 2.59.
There is some evidence that cardiac failure is different in RA. Davis et al. compared 103 RA patients with heart failure with 852 non-RA patients who also had heart failure. RA patients were less frequently obese. Patients with RA and heart failure also had fewer typical symptoms and signs and were less likely to have echocardiography compared with non-RA patients. One-year mortality after heart failure was higher in patients with RA compared with non-RA patients (35% vs. 19%).
Cardiac risk factors
One crucial question is whether the increased cardiovascular morbidity and mortality in RA are due to a higher prevalence of traditional cardiovascular risk factors. The evidence for this possibility has been evaluated in a systematic review by Boyer et al. . They found 15 case-control studies involving 2956 RA patients and 3713 controls. They found strong evidence that the prevalence of smoking was increased in RA patients (odds ratio 1.56 (95% CI 1.35, 1.80)). More RA patient had diabetes mellitus with an odds ratio of 1.74 (95% CI 1.22, 2.50). However, there was no evidence that more RA patients had hypertension or hypercholesterolaemia, although high-density lipoprotein (HDL) cholesterol levels were lower in RA patients with a weighted mean difference of −17 mg dl −1 (95% CI −18, −17), and these may contribute to the increased cardiovascular morbidity of RA.
Treatment effects
Another possible explanation for increased cardiovascular risks is that it reflects treatment effects. There has been most concern about non-steroidal anti-inflammatory drugs (NSAIDs). Trelle et al. have reported the results of a systematic review and network meta-analysis of all large-scale randomised controlled trials involving NSAIDs. The primary outcome was myocardial infarctions. Other outcomes included deaths from cardiovascular disease and deaths from any case. They evaluated 31 trials which had enrolled 116,429 patients and included over 115,000 patient years of follow-up. They found that rofecoxib gave the highest risk of myocardial infarction compared to placebo with a rate ratio of 2.12 (95% CI 1.26, 3.56). Etoricoxib (rate ratio 4.07; 95% CI 1.23, 15.7) and diclofenac (rate ratio 4.07; 95% CI 1.48, 12.7) had the highest risks of cardiovascular deaths. They concluded that NSAIDs have as a group a general increased risk of cardiovascular disease. Naproxen appeared to have the least risks.
Corticosteroid use is also a potential factor increasing influence cardiovascular disease in RA and this risk has been assessed by Ruyssen-Witrand et al. . They identified 37 studies which assessed cardiovascular risk in RA patients treated with low-dose corticosteroids (under 10 mg day −1 prednisone). The analysis showed relative heterogeneous results. There was some evidence for a protective effect on serum lipid profiles but there was an increase of insulin resistance and high blood sugar levels. In four out of six studies, there was evidence of an association with major cardiovascular events. This included myocardial infarctions (HR 1.7; 95% CI 1.2, 2.3) and a composite index of cardiovascular events in rheumatoid factor-positive RA (HR 2.2; 95% CI 1.2, 4.0). On balance, steroids appear to increase cardiovascular risks, although the evidence is variable and incomplete.
Disease-modifying drugs, particularly methotrexate, could also influence cardiovascular disease in RA. The impact of methotrexate has been assessed by Micha et al. . Their systematic review identified 10 observational studies in which the impact of methotrexate on cardiovascular disease in RA was assessed. They found that methotrexate treatment reduced the risk of cardiovascular disease by 21%, including an 18% lower risk for myocardial infarction. They suggested that the effective reduction of RA inflammation by methotrexate is likely to reduce cardiovascular disease risks. Another systematic review of the impact of methotrexate by Westlake et al. drew broadly similar conclusions. They found two studies that specifically assessed the relationship between methotrexate use and cardiovascular mortality. One study showed significant reductions in cardiovascular mortality and the other reported a trend towards reduction. A further five studies considered all-cause cardiovascular morbidity: four showed reductions in cardiovascular morbidity with methotrexate and the fifth study a trend towards reduction. Finally, four studies considered myocardial infarction: one showed a decreased risk and three a trend towards decreased risk with methotrexate.
Finally, tumour necrosis factor (TNF) inhibitors might affect cardiovascular risks. This has been evaluated by Westlake et al. who identified 20 studies that evaluated this question. They found some evidence that TNF inhibitors decrease cardiovascular morbidity in RA. They found no definite association with myocardial infarctions and heart failure, although the number of individual events was small and unlikely to reach statistical significance. In three studies, TNF inhibitor responders had lower risks of cardiovascular events. Most studies showed no relationship between TNF inhibitor use and heart failure.
Cardiovascular risks in long-term inflammatory diseases
RA may not be an exceptional disease with respect to increased cardiovascular disease. There is evidence that chronic inflammatory diseases in general have higher rates of cardiovascular disease. Roifman et al. compared risks in several chronic inflammatory diseases including systemic lupus erythematosus, ankylosing spondylitis, psoriatic arthritis, polymyositis and dermatomyositis, and inflammatory bowel disease as well as RA. In all these diseases, there was some evidence of an increase in cardiovascular diseases. These findings suggest that it is the presence of chronic inflammation rather than RA itself which is the main link, although the exact nature of the inflammatory drivers remains controversial.
Osteoporosis
Periarticular osteoporosis and osteopenia occur in RA, but generalised bone loss is also common, likely due to a shared mechanism via cytokine-induced osteoclast activation and osteoblast inhibition leading to bone resorption ( , reviewed in ). The presence of increased rates of fractures in RA and the associated osteoporosis risk factors have been known for many years . Part of this is due to reduced peripheral bone density in RA patients, which is related both to age and to steroid therapy as well as disease duration , and markers of RA severity including positive rheumatoid factor and radiographic damage . A large North American observational study, which enrolled 8419 female RA patients, showed that increases in fracture risks were associated with post-menopausal status, HAQ scores and prednisone use, while TNF inhibitors were associated with decreased overall fracture risk. Interestingly, although 80% of patients had at least one risk factor for osteoporosis, only 32% were on preventive treatments .
Pye et al. studied the influence of disease-related variables on hand cortical bone loss in women with early RA and other forms of inflammatory arthritis in a study nested within the Norfolk Arthritis Register. They evaluated 204 women and assessed changes in bone mineral density over 4 years, which declined at a rate equivalent to 1% per year. They found that increased disease activity and severity were associated with accelerated bone loss. This suggests that intensive early treatment will limit the severity and progression of bone loss in RA. There is also evidence that the same factors are involved in the risk of vertebral fractures. El Maghraoui et al. studied 172 women with RA, none of whom was taking osteoporosis medications. They found that 36% of patients had evidence of vertebral fractures. These fractures were associated with more long-standing and severe disease and a greater use of corticosteroids.
Periodontal disease
There has been considerable recent interest in the relationship between periodontitis and RA. The two conditions share common risk factors, including smoking and HLA-DR B1 04 alleles as well as shared pathological processes . Periodontitis, which involves inflammatory damage of the periodontal attachment and alveolar bone, is more common in RA than in healthy controls. There is also evidence that periodontitis and other oral infections are involved in RA pathogenesis .
There is strong evidence that patients with established active RA have a higher frequency and severity of periodontal disease compared with healthy subjects. Pischon et al. compared periodontitis in 57 RA patients and 52 age- and sex-matched healthy controls. RA status and age were the only significant predictors of periodontal disease. RA patients had an eightfold increased odds of periodontal disease (95% CI 3, 22). de Pablo and colleagues studied people within the third National Health and Nutrition Examination Survey (NHANES III), a nationally representative cross-sectional survey of non-institutionalised civilians from North America. They investigated a sample of 4461 participants; 103 of these had RA. The participants with RA had more missing teeth (20 vs. 16 teeth) compared to participants without RA. After adjusting for age, sex, ethnicity and smoking status RA patients were more likely to have lost all their teeth (odds ratio 2.3; 95% CI 1.6, 3.3) and to have periodontitis (odds ratio 1.8; 95% CI 1.0, 3.2) compared with non-RA subjects. Seropositive RA was particularly associated with edentulism.
Gastrointestinal disorders
RA is associated with a range of gastrointestinal (GI) co-morbidities . Some of these are associated with the disease itself, but most are related to its treatment. NSAIDs in particular are a major risk factor for a range of bowel and liver problems. It is challenging to consider what exactly constitutes a GI co-morbidity because some features of RA, such as systemic amyloidosis, can rarely affect the GI tract, although their effects are widespread. Other disorders, such as Sjogren’s syndrome, can have indirect effects on the GI tract, although they are essentially disorders of other organs. The main GI co-morbidities of RA affect the upper GI tract, the small bowel, the large bowel or the liver. They are summarised in Table 1 .
| Type | Frequency | Related to treatment | Main features | Key investigations |
|---|---|---|---|---|
| Oesophagitis | Common | NSAIDs | Dyspepsia | Endoscopy |
| Gastro-duodenal ulceration | Common | NSAIDs and Corticosteroids | Dyspepsia and anaemia | Endoscopy |
| Small bowel disease | Relative common | NSAIDs | Anaemia and weight loss | Capsule enteroscopy |
| Colitis | Uncommon | NSAIDs | Diarrhoea | Colonoscopy |
| Hepatitis | Uncommon | NSAIDs and DMARDs | Usually asymptomatic biochemical findings | Biochemical tests tests and non-invasive imaging (eg fibroscan) |
Upper GI disease
Oesophagitis, gastric ulcer disease and duodenal ulcer disease are a continuum, and in the main they are related to NSAID prescribing for RA. They can variously lead to mild symptoms such as dyspepsia, GI bleeding or severe problems such as perforations. Historically, this was a major problem. Over 20 years ago Fries et al. found NSAID-induced upper GI disease a major problem. Patients on NSAIDs had an HR for GI hospitalisation that was more than 6 times that of patients not on NSAIDs. They estimated that the syndrome of NSAID-associated gastropathy accounted for 2600 deaths and 20,000 hospitalisations each year for RA patients in the USA.
However, the situation has improved over time. A subsequent study by Fries et al. , which reported experience in 5598 RA patients followed for 31,262 patient-years at eight sites, showed temporal changes in NSAID gastropathy. Rates for GI-related hospital admissions rose initially from 0.6% in 1981 to 1.5% in 1992. However, they then declined to 0.5% by 2000. This decline was related with lower doses of ibuprofen and aspirin, a decline in the use of ‘more toxic’ NSAIDs, a rise in the use of ‘safer’ NSAIDs and increasing use of proton-pump inhibitors. A recent Dutch study by Steen et al. reported similar findings. In 1997, in Amsterdam, the incidence of symptomatic GI events was 2.1% in RA patients. A subsequent prospective survey in 2003 showed a fall to 1.2%. They too suggested that the decline of NSAID-induced GI events was most likely due to both greater adherence to guidelines for prevention of NSAID gastropathy and better overall RA treatment.
Corticosteroids are also considered to be implicated in upper GI disease in RA, Although the evidence is less convincing than for NSAIDs and rather more historical . In part, the impact of steroids may be most marked with severe risks such as upper GI perforation. Curtis et al. evaluated 37 hospital admissions for GI perforation from 40 841 RA patients, of which 37 hospitalisations with GI perforation were identified. Seventy percent of patients with GI perforation had received steroids. The HR was particularly high in patients who were receiving both steroids and NSAIDs (HR 4.7; 95% CI 1.9, 12.0).
Small-bowel disease
NSAIDs also cause small-bowel disease. Increased intestinal permeability is considered to be the central mechanism which translates biochemical damage to tissue damage. NSAIDs increase intestinal permeability almost as soon as they are taken. It results in significant morbidity and occasionally mortality . The clinical features include significant bleeding, weight loss, perforation, obstruction and even sudden death . Capsule endoscopy is the best way to investigate small-bowel enteropathy. Observational studies show that over 60% of patients receiving NSAIDs had abnormal features. The main changes were reddened folds, denuded areas and mucosal breaks. Two percent had diaphragm-like strictures and 3% had bleeding without an identifiable lesion.
Colitis and large-bowel disease
NSAIDs can also cause damage to the colon, although this appears to be relatively uncommon. NSAID use has been associated with a range of colonic problems including bleeding, iron deficiency anaemia, strictures, ulcerations, perforations, diarrhoea and death . NSAIDS may also exacerbate pre-existing inflammatory bowel disease . Bowel perforation occurs rarely, associated with glucocorticoids and the presence of diverticulitis . More recently, GI perforation has been described in patients receiving tocilizumab therapy , although rates were lower than in patients receiving corticosteroids .
Hepatitis and liver disease
Liver function tests are often abnormal in RA; they mainly comprise increases of alkaline phosphatase and serum gamma-glutamyltransferase levels . These changes may reflect the acute phase response in active RA rather than indicate significant liver damage.
In a study of 98 patients with RA, 12 showed increases of both alkaline phosphatase and γ-glutamyltransferase levels, but isoenzyme alkaline phosphatase studies disclosed that in three of seven patients, an increase in bone isoenzyme was responsible for the rise of alkaline phosphatase. In the others, the liver isoenzyme alone was raised, and, in one patient, both were raised.
Occasional patients have been reported in which chronic hepatitis was found in association with RA. For example Job-Deslandre et al. reported six patients in whom chronic hepatitis developed after the onset of RA. There were mild increases in transaminase levels. Liver biopsies showed active chronic hepatitis in five cases with cirrhosis in three cases. The evidence for a genuine association remains incomplete.
Disease-modifying anti-rheumatic drugs (DMARDs), including methotrexate and leflunomide, and NSAIDs have all been implicated in causing liver damage in some patients. However, the frequency of these problems is not fully known. For example, Quintin et al. reported a case-control study involving 1571 arthritis patients on long-term low-dose methotrexate therapy. Results of liver biopsies from 41 patients with elevated liver enzymes were compared with 41 disease control subjects also receiving methotrexate who did not have elevated liver enzymes and were matched for age and sex. Liver biopsies showed 17 autoimmune hepatitis-like lesions, 13 non-alcoholic steatohepatitis-like lesions, seven limited liver lesions and two primary biliary cirrhoses. However, methotrexate-specific liver lesions were rarely seen in these patients. When there are major concerns about potential liver toxicity due to DMARDs, non-invasive tests, such as fibroscans, appear most appropriate .
Abnormal liver function tests are seen more often in RA patients receiving methotrexate and leflunomide. Curtis et al. assessed the frequency of elevated liver enzymes in these patients. These elevated levels (over one time the upper limit of normal) occurred in 14–35% of patients receiving different DMARD treatments. The combination of methotrexate and leflunomide was most likely to cause high levels.
Finally, NSAIDs can also cause liver damage. Although relatively uncommon, NSAIDs are one of the most important causes of drug-associated liver disease . Diclofenac is thought to have relatively high risks of liver damage and the level of risk was assessed in a large clinical trial . A total of 17 289 patients received diclofenac for a mean of 18 months in this study. Overall 3% of patients have liver function tests over three times the upper limit of normal, hospital admissions occurred in four patients (0.02%) and two patients (0.01%) had serious liver disease with elevated enzymes.
Related posts:
Precipitating and perpetuating factors of rheumatoid arthritis immunopathology – linking the triad of genetic predisposition, environmental risk factors and autoimmunity to disease pathogenesis
What is the outcome of RA in 2011 and can we predict it?
Assessment of control of rheumatoid arthritis disease activity
Quality of life and the outcome of established rheumatoid arthritis
Anti-TNF therapy
Imaging in rheumatoid arthritis
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
