The current COVID-19 pandemic raises several clinical challenges. Cases of COVID-19-associated arthritis have been reported, and inconsistently described as either COVID-19 viral arthritis or COVID-19 reactive arthritis. We aimed to review all the reported cases of ‘COVID-19-associated arthritis’, which we propose, is a better term to define the entire spectrum of new-onset arthritis believed to be associated with SARS-CoV-2 infection. We performed a systematic literature review using MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews to search for articles published up to 13 December 2021. We included cohort studies, case series and case reports describing patients diagnosed with COVID-19 reactive or viral arthritis by a physician, irrespective of fulfilment of classification criteria. To identify relevant studies, medical subject headings and keywords related to ‘COVID-19/SARS-CoV-2 infection’ and ‘reactive arthritis’ were used. Our search retrieved 419 articles, of which 31 were included in the review. A total of 33 cases were reported in these 31 articles, the majority being adults (28/33=85%) with peripheral joint involvement (26/33=79%). Most of the patients responded well to treatment and the disease was self-limiting. These 33 case reports describe a possible causal relationship between exposure to SARS-CoV-2 and the onset of arthritis. However, since these cases were reported during a pandemic, other aetiologies cannot be fully excluded. The exact mechanism through which SARS-CoV-2 might trigger arthritis is not fully understood and robust epidemiological data to support a causal relationship are still lacking.
- arthritis, reactive
- arthritis, infectious
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
Cases of SARS-CoV-2-associated arthritis have been reported, and inconsistently described as cases of either viral arthritis or reactive arthritis.
WHAT THIS STUDY ADDS
We performed a systematic literature review of COVID-19-associated arthritis and comprehensively describe 33 case reports, gathering all the information about this potential new entity in one single article.
We propose a unified designation: ‘COVID-19-associated arthritis’—in which SARS-CoV-2 infection has been proven, there is a plausible temporal relation between COVID-19 and the onset of arthritis, and the arthritis cannot be justified by other aetiologies.
Nonetheless, causality between SARS-CoV-2 infection and the onset of arthritis still needs to be proven, as robust mechanistic and epidemiological data are still lacking.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE AND, OR POLICY
COVID-19-associated arthritis is a potential new disease entity, in which varied joint involvement has been reported, generally with a self-limiting course and good response to treatment (non-steroid anti-inflammatory drugs and glucocorticoids).
Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is a novel coronavirus that was first identified in Wuhan, central China, and is responsible for the 2019–2022 pandemic. Globally, there have been 512 million confirmed cases of Coronavirus disease 2019 (COVID-19), the disease caused by SARS-CoV-2, including >6 million deaths, reported to World Health Organization (WHO) up to 4 May 2022.1
Coronaviruses belong to the family of ribonucleic acid (RNA) viruses, and they are characterised by the presence of crown-like spikes on their surface. Hundreds of coronaviruses circulate among animals (eg, cats, pigs, camels and bats) but cross-species virus transmission (‘spill over events’) can occur when the exposure to other species increases and the natural barriers to infection of new hosts are overcome.2 Seven coronaviruses are known to infect humans, including SARS-CoV-2 (table 1).3
Organ dysfunction caused by SARS-CoV-2 can be a consequence of cytotoxic damage (direct injury) and immunological insult (indirect damage) to host cells. Pneumonia and acute respiratory distress syndrome are the major complications of COVID-19. Other COVID-19 complications may include acute liver, cardiac and kidney injury, as well as secondary infection, immunothrombosis and hyperinflammatory response.4
As with other respiratory tract infections, musculoskeletal symptoms can develop with coronavirus infections. Arthralgia is reported in 15% of patients with COVID-19, and myalgia is even more common (44%). However, musculoskeletal symptoms do not seem to be related to COVID-19 severity.5
The current COVID-19 pandemic raises several intriguing questions, as well as clinical challenges in the context of rheumatic and musculoskeletal diseases. Cases of ‘SARS-CoV-2-associated arthritis’ have been reported, and inconsistently described as cases of either viral arthritis or reactive arthritis. This distinction is sometimes difficult, and classically the term reactive arthritis has been used to describe an immune-mediated sterile arthritis (typically monoarticular or oligoarticular, often involving the lower extremities, and sometimes associated with dactylitis and enthesitis) occurring several days to weeks after bacterial infection of the gastrointestinal or genitourinary tract, in genetically predisposed individuals; while the term viral arthritis has been used to describe the (often symmetric and polyarticular) articular manifestations of viral infections, potentially triggered by a variety of mechanisms, including direct invasion of the joint cells or tissues, immune complex formation and direct or indirect immune dysregulation due to persistent/latent viral infection, sometimes in the context of an upper respiratory tract infection. Microorganisms that have been associated with reactive and viral arthritis are listed in table 2.6
To date, there is no agreement regarding viral or reactive arthritis diagnostic or classification criteria. The diagnosis is essentially clinical, based on a careful history and physical examination.7 In this article, we aimed to review all the reported cases of ‘COVID-19-associated arthritis’, which we propose, is a better term to define the entire spectrum of new-onset arthritis believed to be associated with SARS-CoV-2 infection.
Protocol and search strategy
The systematic literature review (SLR) was conducted and reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.8
The literature search was performed in MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews (CENTRAL) for articles published up to 13 December 2021. We selected published articles in English, Portuguese, Turkish, Spanish or French. To identify the relevant studies, medical subject headings and keywords related to ‘COVID-19/SARS-CoV-2 infection’ and ‘reactive arthritis’ were used. The search strategies used to identify relevant studies are provided in online supplemental table 1.
The American College of Rheumatology and European Alliance of Associations for Rheumatology (EULAR) meeting abstracts from the last 2 years (2020–2021) were searched and only included if they had not already been published as original studies. Reference lists of all relevant studies retrieved from the electronic search were manually searched to identify additional potentially eligible studies.
The following study designs were allowed: cohort studies, case series and case reports; to gather information about the clinical, biological and demographic features of these patients.
EndNote V.20 was used to manage the references obtained from the search results of each of the databases. The protocol was written and defined before starting the search and no deviations were performed during the process.
Two reviewers (ASP and BF) independently screened the titles and abstracts of the retrieved articles applying predefined inclusion criterion: patients diagnosed clinically with COVID-19 reactive or viral arthritis by a physician, irrespective of fulfilment of published classification criteria for these conditions. We excluded articles with incomplete information, non-human studies, study population/individuals with pre-existing chronic inflammatory rheumatic diseases or in which another rheumatic disease was subsequently diagnosed and could be the cause of the new-onset arthritis, cases of arthritis associated with vaccination against SARS-CoV-2, and other types of studies not specified above (eg, reviews without original data and editorials).
Titles, abstracts and full texts were screened for inclusion by two independent authors (ASP and BF). Any discrepancies were discussed until a consensus was reached or with the involvement of a third reviewer (PMM) whenever necessary.
Data were collected from each article in a data collection sheet and included: author(s), country(ies) of origin of the data, year of publication, study design, number of patients, demographic features (age, gender), full details about SARS-CoV-2 infection diagnosis and symptoms, tests used to confirm the diagnosis, imaging investigations and medications. Other data included: time of presentation of arthritis, time between SARS-CoV-2 infection/COVID-19 onset and articular symptoms, site and number of joints involved and joint aspiration results. Serological tests were included if they were performed, as well as immunological tests including rheumatoid factor and anticitrullinated protein antibodies, and other laboratory tests such as acute phase reactants and full blood count results. The chronicity of the manifestations was recorded.
Data were summarised by stratifying the results into two groups: COVID-19-associated arthritis with onset up to 1 week after the initial COVID-19 manifestations, and COVID-19-associated arthritis with onset >1 week after the initial COVID-19 manifestations. The 1-week cut-off was arbitrarily defined, in order to distinguish early onset cases from medium-onset/late-onset cases, as they may have different underlying pathophysiological mechanisms.
Our search retrieved 419 articles (414 identified via searching electronic databases and 5 identified from citation searching). Of these, we excluded duplicates within and across databases, and a total of 346 articles were assessed (341 from electronic databases and 5 from hand search). After review of the titles and abstracts, 46 articles were retrieved for full-text evaluation, of which 31 were included in the review (26 from electronic databases and 5 from hand search). A total of 33 cases were reported in these 31 articles. Figure 1 depicts the flow of information through the different phases of the SLR. The 15 articles excluded and the reasons for exclusion are shown in online supplemental table 2.
COVID-19-associated arthritis with onset up to 1 week after COVID-19
A total of six articles reported seven cases of onset of arthritis up to 1 week after COVID-19 diagnosis (table 3). Of these, three cases occurred in children9 10 and four in adults.11–14 Onset of arthritis was 1 week before COVID-19 symptoms in one case,14 concomitant in one case11 and within the first week of onset of COVID-19 symptoms in the remaining five cases.9 10 12 13 All the cases reported had a confirmation of the infection by nasopharyngeal and/or oropharyngeal swab by real-time-polymerase chain reaction (RT-PCR), except one, in which infection was confirmed by a positive IgM against SARS-CoV-2 detected by enzyme linked immunosorbent assay (ELISA).10 Only one case was Human Leukocyte Antigen (HLA)-B27 positive.12
Polyarthritis was observed in three cases,11 12 14 oligoarthritis in two cases9 10 and monarthritis in two cases.10 13 In the report by Liew et al,13 balanitis (a typical extra-articular manifestation of classical reactive arthritis) was observed in association with knee monarthritis.
While non-steroid anti-inflammatory drugs (NSAIDs) were used in two paediatric patients,10 no specific treatment was used in the third paediatric patient9 and none of them progressed to chronicity of the arthritis. Alivernini et al11 reported that NSAIDs, baricitinib and glucocorticoids were started, Apaydın et al12 reported that hydroxychloroquine (HCQ), glucocorticoids and sulfasalazine were started, Liew et al13 reported that NSAIDs were started and intra-articular triamcinolone was administered in the knee joint and Talarico et al14 reported that medium doses of glucocorticoids were started. There was one case in which the patient did not achieve remission before publication of the article12 and the remaining where chronicity was not reported.
COVID-19-associated arthritis with onset >1 week after COVID-19
A total of 25 articles reported 26 cases of onset of arthritis >1 week after COVID-19 (table 4). Except for two patients aged 14 and 16 years,15 16 all were adult patients, and the ages ranged from 27 years17 18 to 73 years.19 Half of the patients (13/26) were male.15 17 19–29 The time of occurrence of arthritis after COVID-19 varied between cases. One report did not specify the period of time between onset of COVID-19 and onset of arthritis, describing it as a ‘short period’,20 while for the remaining cases, the shortest period of time was 10 days,30 31 and the longest period of time was 48 days (development of polyarthritis 3 days after a 45-day hospitalisation for COVID-19 pneumonia).28 SARS-CoV-2 infection was confirmed by RT-PCR or antigen test in all the patients with exception of one,20 in which a subsequent SARS-CoV-2 antibody test was positive.
The number and location of joints affected were variable. Eight cases were reported as monarthritis17 21 22 24 28 29 32 33; of these, four affecting the knee,24 28 29 32 two the ankle,22 33 one the elbow21 and one affecting the MCP joint.17 In addition to monarthritis, psoriasis was reported in one case21 and Achilles tendinitis in another case.33 Oligoarthritis was reported in three cases; knee and ankle involvement was reported in two cases,23 26 and bilateral ankle arthritis and Achilles enthesitis in one case.27 Polyarticular involvement was reported in eight cases; one of the cases was juvenile15 and seven were adult cases.18 19 25 30 31 34 35 Although the joint involvement pattern differed between cases, proximal and distal interphalangeal joints of the hands and feet, knees, wrists, ankles and elbows were the most frequently affected joints.
In addition to peripheral joint arthritis, sacroiliitis was reported in four cases (one patient with concomitant arthritis of the left first costovertebral and costotransverse joints, one with concomitant monarthritis of the hip and two with isolated sacroiliitis).20 36 37 Extensor tendinitis of the hand was reported in one case,38 and dactylitis with extra-articular involvement (conjunctivitis, oral lesions, palatal erosion, blennorrhagic keratoderma, subungual hyperkeratosis and onycholysis, circinate vulvitis)39 and without other symptoms16 in two cases. Three cases were HLA-B27 positive.20 30 39
NSAIDs and glucocorticoids were the most prescribed drugs for the treatment of the arthritis. NSAID monotherapy was used in nine patients,16 19 21 23 32 33 35 37 glucocorticoid monotherapy in five patients,22 28 30 31 39 NSAIDs plus glucocorticoids in eight patients,15 17 18 20 25 26 34 36 NSAIDs plus intra-articular steroid injection in two patients24 27 and NSAID gel plus opioid drugs in one patient.38 The arthritis resolved spontaneously (without any specific treatment for the arthritis) in one patient.29 '‘Significant improvement’, ‘remission’, ‘regression’ or ‘resolution’ of arthritis was observed in 20 patients,15 16 19–26 28–32 34–36 39 ‘moderate improvement’ in 1 patient,27 ‘residual swelling’ in 1 patient,17 ‘tenderness’ in 1 patient,38 low back pain in 2 patients37 and persistent synovitis (‘synovitis still present’) was reported in 1 patient,33 at the last evaluation of the cases. Table 5 presents a summary of patients’ characteristics and their frequency among reported cases.
In this systematic review, we gathered all published data on patients with COVID-19-associated arthritis. We summarised a total of 33 cases in two categories: onset up to 1 week after COVID-19 (7 cases), and onset >1 week after COVID-19 (26 cases).
We propose a new classification for cases of arthritis possibly associated with SARS-CoV-2 infection: ‘COVID-19-associated arthritis’. The distinction between reactive and viral arthritis in the context of SARS-CoV-2 infection is artificial and there is no consensual definition for either of them, both in terms of clinical presentation as well as in terms of the required period of time between the onset of COVID-19 and the onset of arthritis. Some of the cases described are outside of the interval of time described as typical of reactive arthritis (1–4 weeks); these are reports in which the articular symptoms started before or at the same time as COVID-19 symptoms or in which the articular symptoms started at least 1 month after of the diagnosis of COVID-19.
Since important differential diagnoses of viral and reactive arthritis include crystal arthropathy, septic arthritis and other chronic inflammatory joint diseases,40 serological evaluation and synovial fluid analysis are critical investigations. Arthrocentesis results were reported in 10 cases,11–13 21 23 26–29 32 and no crystal and/or microorganisms were detected in the synovial fluid of these patients.
Our study has limitations. It is based on a small number of reports and these being descriptive and voluntary observations they are prone to ascertainment and publication bias, and findings may not be generalisable to other populations and settings. Moreover, some cases were better documented than others, and for example, important investigations such as autoantibody and serological testing and synovial fluid analysis were not systematically performed. Furthermore, the SLR did not find epidemiological data about the incidence or prevalence of cases of viral and/or reactive arthritis during specific periods of the pandemic, which could be compared with the incidence or prevalence of cases of COVID-19 infection during the same period of the pandemic, and with the incidence or prevalence of cases of viral and/or reactive arthritis before the pandemic, in the same geographical area—this type of data could allow to better infer regarding a potential causal association between COVID-19 and arthritis, as studied in Guillain-Barré syndrome.41
Despite these limitations, taken together, these case reports describe a possible causal association between exposure to SARS-CoV-2 and the onset of arthritis. All the cases reported had a confirmed diagnosis of COVID-19, defined by either a positive nasal or throat swab RT-PCR for viral RNA or a subsequent positive serological test for anti-SARS-CoV-2 IgM or IgG. There was only one case in which a nasal or oropharyngeal swab was not obtained during the symptomatic period,20 but the patient symptoms were highly suggestive of COVID-19, and a subsequent SARS-CoV-2 antibody test (IgG) was positive. However, since these cases were reported during a pandemic period, we cannot exclude that they could simply be the reflection of the background incident cases of viral and reactive arthritis, for which the list of potential culprits is extensive (table 2) and often undetermined in clinical practice. Since we do not have a full picture of the incidence of viral and reactive arthritis during the pandemic period that could be compared with prepandemic figures (in the same geographical area), in which SARS-CoV-2 infection was not an existing cause for arthritis, we cannot be sure that these cases are truly associated with SARS-CoV-2 infection and exclude the possibility of coincidental cases (caused by other virus or bacteria). Nevertheless, in most cases, the authors concluded that the most likely diagnosis was viral or reactive arthritis associated with COVID-19, as other causes to explain the arthritis were excluded by autoantibody testing/serology, arthrocentesis and patient history/symptoms. Conversely, it could also be argued that the reported diagnostic workup was not sufficiently extensive and was not done consistently and in a standardised way across reports.
Furthermore, the exact mechanism through which SARS-CoV-2 might cause arthritis is not fully understood, and mechanistic data are still lacking. The most common hypothesis is the existence of molecular mimicry between SARS-CoV-2 viral epitopes and the synovial membrane causing local inflammation, but other theories have proposed a role for the presence of circulating immune complexes or localisation of the virus directly on joint tissue.42 Molecular mimicry triggers humoral and cellular autoreactivity in the host.43 Primary SARS-CoV-2 infection induces systemic inflammation that can impact the musculoskeletal system allowing direct viral infection.42 Recent studies have found an association between the microbiome and reactive arthritis. Any disruption of the equilibrium between gut and microbiome is called dysbiosis, which sometimes can activate a T helper 17-mediated immune response in the host’s intestinal lamina propria, thereby promoting local and systemic inflammation,44 which could be another pathogenic link between SARS-CoV-2 and arthritis. The dysbiotic gut microbiota that persists after disease resolution could be a factor predisposing to the development of persistent symptoms and/or multisystem inflammation syndromes that occur in some patients after clearing the virus.45
In conclusion, this study summarised 33 cases of (possible) COVID-19-associated arthritis. The pattern of joint involvement described in these cases was diverse, but most patients had peripheral involvement with either polyarthritis or monarthritis. Most of the patients responded well to treatment (NSAIDs and glucocorticoids).
Mild disease symptoms resembling viral or reactive arthritis, a plausible temporal relationship between SARS-CoV-2 infection and the onset of arthritis, good response to treatment and low propensity to chronicity of the arthritis are preliminary characteristic features of COVID-19-associated arthritis. However, extensively excluding other possible causes of arthritis remains critical and challenging. Therefore, data should be interpreted with caution and bearing in mind the lack of robust mechanistic and epidemiological data, and that cases were reported with a short follow-up time after the onset of arthritis.
Moreover, we feel that the term ‘COVID-19-associated arthritis’ is more appropriate because it acknowledges the unknown mechanistic links, the lack of epidemiological data, the heterogeneity of clinical presentation and the lack of uniformity in the literature, with these cases ‘randomly’ being named as either reactive arthritis or viral arthritis. Homogenising the nomenclature will hopefully contribute to more consistent data collection and reporting, and further investigation of this new potential condition. As knowledge evolves, refinement of the nomenclature and/or establishment of robust classification/diagnostic criteria is expected to occur.
The definitions of viral and reactive arthritis (in general) are still evolving and the inconsistency in nomenclature found in this review suggests that the use of the term ‘COVID-19-associated arthritis’ is more appropriate and inclusive, when describing arthritis in the context of SARS-CoV-2 infection, at least until more mechanistic data regarding the potential link between SARS-CoV-2 and arthritis are available. Our review informs regarding the interpretation of future case reports of COVID-19-associated arthritis and the refinement of the definition and characteristics of this potential emerging new entity.
Patient consent for publication
BF and ASP contributed equally.
Contributors BF and ASP performed the literature search, performed the data extraction and analysis and wrote the first draft of the manuscript. PMM designed the study, supervised the work and acted as systematic literature review methodologist and third reviewer. All the authors contributed to writing the manuscript, read and approved the final manuscript.
Funding BF is supported by the Turkish Society for Rheumatology. ASP is supported by a scientific training bursary for young fellows from EULAR. PMM is supported by the National Institute for Health Research (NIHR), University College London Hospitals (UCLH), Biomedical Research Centre (BRC).
Disclaimer None of the institutions had a role in the study, including study design, collection, analysis and interpretation of data; writing of the report and decision to submit the paper for publication.
Competing interests PMM has received consulting/speaker’s fees from AbbVie, BMS, Celgene, Eli Lilly, Galapagos, Janssen, MSD, Novartis, Orphazyme, Pfizer, Roche and UCB, all unrelated to this manuscript. ASP and BF: no conflicts of interest.
Provenance and peer review Commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.