State of the art: what we know about infectious agents and myositis

Purpose of reviewIncreasing evidence suggests that the idiopathic inflammatory myopathies (IIMs) result from certain environmental exposures in genetically susceptible individuals. Investigations have demonstrated that a variety of infections not only cause infectious myopathies but also could be possible triggers for IIM. This review summarizes published studies on the possible roles of infections in inflammatory muscle disease. Recent findingsMany infectious agents have been linked to the development of IIMs via case reports, epidemiologic investigations, and animal models. Additional agents possibly involved in triggering the development of IIMs have been recently described, including Torque teno virus (TTV) and Borrelia burgdorferi. Novel animal models of myositis have been recently developed using Leishmania infantum or Chikungunya virus (CHIKV). New technologies to assess infectious agents include high-throughput methods for pathogen identification and novel approaches to identify gene expression of pathogens in tissues. SummaryUnderstanding the causes of IIMs remains limited in part due to the rarity and heterogeneity of these disorders. Although no definitive studies have yet linked infectious agents with IIMs, additional evidence is accumulating and novel technologies may allow improved understanding of the roles of infections in IIMs and for possible future therapeutic and preventive measures.


Introduction
Inflammatory myopathies, defined by chronic inflammation in muscles, are characterized clinically by a wide variety of symptoms including muscle tenderness, weakness, swelling, and pain, and include a large number of conditions such as infectious, toxic, endocrine, and other myopathies. Significant advances in defining the pathologic and molecular features of the inflammatory myopathies have been made recently [1,2,3 ], yet much remains unknown. When clinical, laboratory, and pathologic studies fail to identify the known causes of inflammation in a muscle, a diagnosis of idiopathic inflammatory myopathy (IIM) may be made. The most common IIM subtypes are dermatomyositis, polymyositis, and inclusion body myositis (IBM). The phenotypic heterogeneity observed in IIM may be related to heterogeneity in genetic and environmental risk factors [4,5], suggesting that the difficulty in understanding the role of infections in IIM may in part relate to the fact that different phenotypes have different infectious risk factors. In this review, we summarize recent findings on associations between infectious agents and onset of IIM and review previous reported pathogens related to infectious myopathies. Our review focuses on articles published in the last 10 years and we do not discuss infections occurring after the diagnosis and therapy of IIM, although these opportunistic infections are common and have been well documented [6].

General issues relating to infections and myositis
Because a broad spectrum of infectious agents -including viruses, bacteria, fungi, and parasites -can infect muscle and cause a primary inflammatory response [7][8][9], it can be difficult to distinguish between true infectious myopathy and what may be infectious causes of what are currently defined as typical IIM cases. Primary ways to make such distinctions often rely on the specific clinical features of the disease, the tempo with which the disease develops after an infection, and whether the disease resolves with antiinfective therapy. It is also useful to consider certain criteria that have been proposed to assess Purpose of review Increasing evidence suggests that the idiopathic inflammatory myopathies (IIMs) result from certain environmental exposures in genetically susceptible individuals. Investigations have demonstrated that a variety of infections not only cause infectious myopathies but also could be possible triggers for IIM. This review summarizes published studies on the possible roles of infections in inflammatory muscle disease. Recent findings Many infectious agents have been linked to the development of IIMs via case reports, epidemiologic investigations, and animal models. Additional agents possibly involved in triggering the development of IIMs have been recently described, including Torque teno virus (TTV) and Borrelia burgdorferi. Novel animal models of myositis have been recently developed using Leishmania infantum or Chikungunya virus (CHIKV). New technologies to assess infectious agents include high-throughput methods for pathogen identification and novel approaches to identify gene expression of pathogens in tissues.

Summary
Understanding the causes of IIMs remains limited in part due to the rarity and heterogeneity of these disorders. Although no definitive studies have yet linked infectious agents with IIMs, additional evidence is accumulating and novel technologies may allow improved understanding of the roles of infections in IIMs and for possible future therapeutic and preventive measures. the role of environmental agents in the development of autoimmune diseases, including the appropriate temporal association, lack of alternative explanations, improvement of the condition after removing the agent or worsening upon re-exposure, biologic plausibility, analogy, dose responsiveness, and specificity [10]. Muscle or other tissue disorder is often useful here as well, as true infectious myositis would be expected to show primarily phagocytes and neutrophils, whereas IIM muscle biopsies more commonly reveal mononuclear infiltrates. Another complicating factor is that the immune changes associated with IIM, which likely develop long before clinical manifestations, may in turn alter the likelihood of infections and create the appearance of an association with that infection when it actually is a secondary event. An alternative theory that has not been adequately explored is that exposure to infections may actually decrease the likelihood of autoimmune diseases [11].
As many autoimmune diseases have been previously associated with infectious agents, investigators have considered infections as possible triggers for IIM as well. It is unclear when the concept that infections may be a cause of IIM was first proposed. In 1936, two cases of dermatomyositis were associated with bacterial or viral infections [12]. Over the years, many additional agents have been implicated with IIM or found to induce inflammation in muscle that have resulted in a large array of infections under consideration (Table 1).
Some studies suggest an infectious cause of IIM based on evaluation of exposures to infections of many types prior to the development of disease. A recent study [13 ] evaluated 285 children with juvenile IIM (JIIM) and carefully reviewed the recorded environmental exposures within 6 months prior to disease onset and found that infections accounted for 44% of the documen-ted environmental exposures. Another two studies evaluated the roles of infections prior to development of juvenile dermatomyositis (JDM). One group in Canada studied 110 patients and identified clinical indications of infection 3 months prior to JDM onset in 71% of cases [14]; another group in the USA evaluated medical records of 286 patients and found frequent complaints of respiratory or gastrointestinal infection symptoms 3 months before the diagnosis of JDM, which resulted in 63% of these cases being given antibiotics [15]. In addition to studies on general infections, case-control studies, case reports, and case series of a specific pathogen have also been published. We discuss these below in four categories: viral, bacterial, fungal, and parasitic infections and summarize the published studies since 2000 in Table 2.

Viruses and idiopathic inflammatory myopathy
The association between viral infections and onset of IIM has a longer history and the evidence appears stronger than that for other pathogens. Epstein-Barr virus (EBV) is a human herpesvirus that resides in a latent form in memory B cells in the majority of the world population and has been reported to be related to the development of a number of autoimmune disorders [48]. A casecontrol study [16 ] showed evidence for higher frequencies of anti-Epstein-Barr nuclear antigen 1 (EBNA1) antibodies at the onset of dermatomyositis/polymyositis and the EBV genome was detected in a higher frequency of patients than in the matched healthy controls. In this study, the concurrence of malignancies such as nasopharyngeal carcinoma (NPC) further increased the risk of development of IIM. In addition, a case report from Japan described the development of polymyositis after EBV infection [17]. Interestingly, a case report of EBV infection was associated with the development of concurrent JDM and type 1 diabetes [18]. Another study [19], however, did not replicate these findings in JDM patients.
Hepatitis viruses and their vaccines have been proposed to be involved in the development of IIM. Case reports suggest a possible association between exposure to hepatitis B virus and the onset of IIM [20], as well as a possible association with hepatitis C virus infection [21][22][23]. In addition to the direct viral infection detected in IIM patients, indirect evidence of the role of viral vaccine-related IIM was also discussed in some previous case reports [49,50].
Another category of viruses, retroviruses, which include HIV and human T-lymphotropic virus 1 (HTLV-1), has been related to development of IIM in case reports [24,27]. Thirteen patients were reported to develop polymyositis after HIV infection, with an average time of 4.3 years from viral infection to disease onset [26] and four patients who developed IBM after exposure to HIV were also studied [25]. A study of 11 Japanese patients with IBM [28] and a report of three British patients with polymyositis [29] both detected anti-HTLV-1 antibodies in sera of all the patients. Additionally, two separate studies of polymyositis patients from Jamaica reported 63% [30] and 87.5% [31] as having serological evidence of prior HTLV-1 infection, respectively. Detection of the viral genome in muscle biopsies further supported the association between HTLV-1 and polymyositis [31].
Although there is evidence that parvovirus B19 infection may be related to certain autoimmune disorders such as rheumatoid arthritis [51], less convincing data exist for the role of parvovirus B19 in the development of IIM. Case reports, including that of an adult with dermatomyositis [32] and two children with JDM [33,34], have suggested an association. Nonetheless, a carefully conducted case-control study [35] did not find an increased prevalence of antiparvovirus B19 IgG in the plasma of patients with JDM and a low level of viral DNA was detected in the patients' muscle tissues. Another negative result was found in a case series of seven patients with polymyositis/dermatomyositis [36].
Enteroviruses, which include coxsackieviruses (group A and B) and enteric cytopathogenic human orphan (ECHO) viruses, are potential candidates for inducing IIM based on previous findings of enteroviral RNA in muscle biopsies [37]. However, some studies failed to detect enteroviral RNAs in muscle biopsy samples of JDM patients [38] or coxsackievirus B (CVB) genomes in IIM patients [39]. Similarly, in a case-control study [40] of new-onset JDM, elevated antibody titers for enteroviruses were detected at the same rates in both patients and healthy controls, suggesting that enteroviral infection might simply be a common environmental exposure rather than a trigger for JDM.
Recently, a case-control study [41] in Hungary detected the infection of a novel virus, Torque teno virus (TTV), in both IIM patients and healthy controls. There was no obvious association between the viral infection and the development of IIM, but the higher frequency of TTV infection in severe cases than in mild ones suggested that this viral infection may lead to a more severe IIM.
Influenza virus was previously reported to be associated with polymyositis [52] and dermatomyositis [53]. Cases of dermatomyositis have also been reported following influenza vaccines, raising the possibility that immune responses to antigens shared by the virus and vaccine could be implicated in the development of myositis [54,55].
Bacteria, fungi, parasites, and idiopathic inflammatory myopathy There are well documented records of bacterial and fungal infection as causes for infectious myopathy, yet reports of their role as IIM triggers are relatively few. An early case-control study [42] found evidence of more frequent streptococcal infection in JDM patients than in matched controls. In a case series from Mexico, 30 patients who developed various systemic rheumatic diseases after Mycobacterium tuberculosis infection were studied and five were identified with polymyositis or dermatomyositis [43]. Another possible inducer of IIM is Borrelia burgdorferi, the bacterial agent of Lyme disease. A case report [44] describing the development of dermatomyositis after B. burgdorferi infection proposed this association. Supporting evidence for fungal infections as a trigger for IIM is limited. An earlier case report [45] described a rare case of a combination of myasthenia gravis, pemphigus vulgaris, and Candida albicans infection in a patient who later developed polymyositis and myocarditis. As for parasitic infections, a case report [46] discussed the development of IIM in one of two siblings infected with Toxoplasma gondii. Rebolledo et al. [47] extensively evaluated the advantages of Caenorhabditis elegans as a nematode model for IBM. These advantages include the overall similarities in metabolic processes and muscle structures; the small size, short lifespan, and simple culture techniques for nematodes; the relative simplicity of genetic manipulations; and the ability to create nematodes transgenic for amyloid precursor protein that form intramuscular amyloid aggregates.

Animal models and inflammatory myopathy
Further evidence for the role of infections in inflammatory myopathy comes from animal models ( Table 3). Some of the first animal models of myositis involved encephalomyocarditis (EMC) 221A virus [56], type D retrovirus [57], and coxsackievirus B1 (CVB1) [58]. During the last 10 years, many other infectious models for inflammatory myopathies have been developed. Examples include murine myositis models induced by Trypanosoma cruzi [59,60], Ross river virus [61,62], or CHIKV [63 ]. Syrian hamsters infected with Leishmania infantum also develop one phenotype of myositis similar to polymyositis [64 ]. These experimental IIM models provide a powerful tool for future etiological and pathogenic studies.

Recent findings on infectious myopathy
Influenza viral infection has been frequently reported to be associated with infectious myopathies, especially in childhood. Children with benign acute childhood myositis were sometimes found to have influenza A virus infections [65][66][67] and some of them were detected with the new pandemic influenza A (H1N1) virus in respiratory samples [66,67]. Additionally, a Swiss group identified five cases with flu-like symptoms before muscle involvement and also reviewed 311 previous cases worldwide that suggested that a subgroup of myopathies is caused by influenza virus A/B [68]. Staphylococcus aureus is the major causative agent of pyomyositis, the most frequently reported bacterial infection of muscles [69]. S. aureus infections range from bacteremia to localized infections including muscle and joint fluid culture [70,71]. Confirmatory evidence of S. aureus infections is the PCR identification of toxin genes encoded by the bacterial strain [71]. Streptococcus pyogenes is a less common infection in muscle and has been recently reported in a case report [72] and a case series [69]. Streptococcus anginosus infection of muscle is likely even less common [73].
In some rare cases, muscle infection by Treponema pallidum has been reported and some patients had secondary syphilis after HIV infection [74,75]. Because fungal infection is closely related to the immune status of patients, fungal myositis is found mainly in immunocompromised patients after HIV infection or organ transplantation and Aspergillus is a common cause in these cases [76][77][78].
Other rarer reports of fungal myositis include Candida tropicalis [79] and Candida albicans [80]. As for parasitic myositis, three patients with Haycocknema perplexum infection were reported in Australia [81]. One case study [82] reported a patient with orbital myositis caused by Spirometra erinaceieuropaei. Another case report [83] from France confirmed the infection of T. gondii in a myositis patient through muscle biopsy assessment and serological testing.

Possible mechanisms by which infections may induce autoimmune disease
Many possible mechanisms for the development of autoimmunity following infection have been proposed, but none has been confirmed by rigorous experimentation. In fact, a continuing debate in this area is whether infectious agents induce or enhance autoimmune disease or whether they rather protect humans and animals from autoimmune diseases. The increasing evidence for a possible protective effect of infectious agents on autoimmune disorders has emerged from a variety of sources including animal models [11]. A recently published review [84 ] summarized the proposed mechanisms of infection-induced autoimmune disease with an emphasis on viral antigens. Although some experimental support does exist for a variety of mechanisms, including molecular mimicry, priming autoreactive immune responses, bystander activation, and epitope spreading, none of these cases has conclusive evidence.

Conclusion
Although evidence from multiple approaches, including epidemiologic studies, case reports, and animal models, support the role of infections in IIM, none of these is conclusive and all require additional investigation [13 ,85,86]. Among all the proposed environmental triggers of IIM, infectious agents have been most widely studied owing to their ability to be definitively identified and to elicit strong immune responses and chronic tissue inflammation, which are hallmarks of IIM. The studies mentioned in this review suggest that certain viruses, bacteria, fungi, and parasites may trigger chronic muscle inflammation, and in some cases autoimmunity, in humans and animals. Experimental evidence has demonstrated the existence of infectious agents in serum and muscle of some patients with IIM, and in some cases, at higher rates than that seen in controls. Nonetheless, these results should be interpreted cautiously, as immune responses to infections likely vary depending on the host genetics, the period of exposure, location of infection, and dose of the infectious agent.
One of the difficulties in deciphering associations between infectious agents and onset of IIM is that environmental and genetic factors are intricately interwoven in the initiation and progression of disease, and it is likely that gene-gene, gene-environment, and environment-environment interactions are playing important roles. Thus, it is possible that multiple environmental agents, either together or in a sequence, may be needed to induce autoimmune responses and synergistic interactions between infectious and noninfectious exposures, as observed in certain malignancies [16 ], may play a role.
Nevertheless, these findings may still provide biomarkers and important directions for the successful diagnosis, evaluation, and therapy of IIM. Although no definitive conclusions can be drawn today regarding the role of infections in the cause of the IIM, current clues clearly suggest that additional studies in this area are warranted. Furthermore, new technologies are also becoming available that could enhance these studies. For example, assessing the presence of viral genomes via arrays of all known viruses may be useful to clarify the exposure of individuals to viral infections that may potentially lead to or exacerbate IIM [87,88 ]. Antiviral antibody arrays are another potentially promising approach [89 ]. Also, DNA microarrays have recently been developed for the detection of other pathogens, such as bacteria and fungi [90 ,91 ]. Future studies should include more carefully designed and powered investigations and improved global participations and collaborations, so as to obtain comparative data from a number of populations more quickly and efficiently. The possibility of identifying infectious agents as triggers for subsets of IIM would have important implications in treatment and may have preventive implications as well.