Rudolph's Pediatrics, 22nd Ed.

CHAPTER 198. Pathogenesis of Inflammatory and Autoimmune Disorders

Kathleen E. Sullivan

A detailed overview of the immune system is provided in Chapter 186. This chapter is focused upon those aspects of inflammation, tolerance, and genetics that are particularly relevant to autoimmune disorders.

The immune system exists to defend against infection and it has developed a beautiful yet complex system to distinguish the “self” from foreign moieties that are considered a threat. The immune system may be broadly divided into the innate immune system and the adaptive immune system. The innate immune system is distinguished by hard-wired programs to distinguish patterns characteristic of pathogens, also known as pathogen-associated molecular patterns (PAMPs). The adaptive immune system consists of T-cell and B-cell functions. T cells and B cells adapt to the environment and undergo a training process to learn to distinguish foreign antigens from self-proteins. As the training process will be distinct for each individual, there is opportunity for errors. Accordingly, most autoimmune disorders are thought to arise from errors in the process of establishing self-tolerance. In contrast, defects in the regulation of innate responses lead to autoinflammatory disorders. The distinction is based on the finding of autoreactive B cells and T cells in autoimmune disorders, as opposed to autoinflammatory diseases, which are associated with considerable inflammatory changes but an absence of autoreactive cells.

INNATE IMMUNE SYSTEM

Neutrophilic infiltrates accompany many autoimmune diseases of childhood. Many of the vasculitides and nearly all types of arthritis are accompanied by a neutrophilic infiltrate. These cells may be extremely destructive, although they are ideally suited for eliminating bacteria. Housed within granules are a multitude of proteolytic enzymes, antimicrobial peptides, and proteins, which are released into the phagosome holding the bacteria. Of relevance to immune and inflammatory diseases is the potential for neutrophils to damage healthy tissue in a bystander phenomenon. Recurrent infections lead to end organ damage as a consequence of neutrophil release of granule material into the tissue. In autoimmune or autoinflammatory disorders, the neutrophils are recruited inappropriately but similarly cause substantial damage to tissues over time.

Tissue macrophages represent one of the main sentinel cells in the body; others are mast cells and dendritic cells. Certain pathogens are resistant to neutrophil killing, such as mycobacteria and fungi. In some autoimmune and autoimflammatory disease, characterized by high levels of TNF-α, macrophages can become activated in a poorly understood process, then aggregate and form granulomas.6,7 All granulomas are comprised of activated macrophages, called epithelioid cells, but not all activated macrophages are found in granulomas. These disorders are effectively treated with TNF-α inhibitors.

Natural killer (NK) cells were originally defined by their ability to kill tumor cells in vitro. NK cells kill their targets by forming a synapse and releasing cytotoxic granules into the synaptic cleft.10 The granules contain granzyme B, which activates apoptosis in the target cell, and perforin, which is thought to act as a pore to facilitate entry of granzyme B. The role of the NK cell is poorly understood in autoimmune disease, but murine models of arthritis suggest that NK cells can downmodulate inflammation.11,12 This may be because NK cells kill antigen-presenting cells as part of the homeostatic process. Patients with systemic juvenile idiopathic arthritis (JIA) have markedly abnormal NK cell function, which may contribute to their propensity for hemophagocytosis, as this process is often driven by cytokines derived from NK cells.

The complement proteins are a set of evolutionarily ancient proteins involved in innate recognition of bacteria and other pathogens (see Chapter 189). Patients with early classical pathway component deficiencies not only suffer from recurrent infections because of the loss of opsonic activity but they are also more prone to developing lupus,14,15 an observation that led to the demonstration of an important role for complement in the clearance of apoptotic cells and the establishment of tolerance.

Regulation of Innate Responses

The contribution of the innate immune system to organ damage characterizing many autoimmune disorders may be visualized directly in the form of the neutrophils and macrophages present in inflamed tissues. Although these cells infrequently bear pathologic responsibility for autoimmune diseases, there is a category of disorders, the autoinflammatory diseases, in which neutrophils are indeed at the root of the problem. Their pathophysiology will be discussed in detail in Chapter 209, but it is useful to describe the outlines here.

The unique monogenic autoinflammatory diseases, familial Mediterranean fever, TNF receptor-associated periodic fever syndrome, mevalonate kinase deficiency, and the CIAS1 autoinflammatory disorders have been helpful in shedding light on the less-well-understood polygenic disorders.35 All of these single-gene disorders have a dysregulated inflammatory pathway involving IL-1 β and IL-18 expression (Fig. 198-1). As a consequence, they are also associated with a compromise in apoptosis, a major regulatory mechanism responsible for reestablishing homeostasis following immune activation. For neutrophils, cell death is ensured as their life expectancy is only 18 to 24 hours. To prevent undesirable perpetuation of inflammation, neutrophil apoptosis accelerates after phagocytosis of bacteria. Most known autoinflammatory diseases have delayed apoptosis, which prolongs the inflammatory process. In the case of the NALP3/CIAS1-associated syndromes, defects in cryopyrin appear to lead to increased cell death. However, the death is not the immunologically cold form seen in apoptosis, but a distinct type of death that engenders more inflammation.36

These disorders demonstrate clearly the critical balance between proinflammatory pathways needed for effective host defense and control mechanisms necessary for limiting damage. Both downregulation of cytokine production and downregulation of the cells themselves are critical, and defects disrupting these processes may be detrimental.

ADAPTIVE IMMUNE SYSTEM

B cells develop in the bone marrow, mature in secondary lymphoid organs, and circulate back to the bone marrow when they become plasma cells. Almost every developmental step is accompanied by checkpoints to ensure that autoreactive B cells do not persist and cause damage. The multiple strategies that B cells have evolved to prevent development of autoreactivity are critical because the recombination events that lead to immunoglobulin molecule generation are random, as likely to lead to autoreactivity as pathogen reactivity. In spite of these strategies, a significant number of autoreactive cells exit into the periphery.16 These cells must compete for T-cell help, and B cells that do not receive T-cell help survive only briefly.17 Generally, the help that B cells require is cognate (ie, the B cell and T cell recognize the same antigen), although this requirement for help from cognate T cells is not absolute. It is believed that during infections, T cells may rescue autoreactive B cells nonspecifically. Similarly, murine models in which B cells have been altered so that they have compromised deletion capability nearly always develop autoimmunity.18,19

FIGURE 198-1. Autoinflammatory disorders affect the inflammosome. The inflammosome is a complex of intracellular proteins that serve to recognize a variety of threats. The inflammosome can be activated by such diverse agents as uric acid, bacterial toxins, bacterial cell wall products, and some viruses. The classic inflammosome is composed of NALP3/CIAS1 (NOD-like receptor family, pyrin domain containing 3), ASC (apoptosis-associated speck-like protein containing a CARD), and procaspase 1. Activation and association of this complex leads to cleavage of procaspase 1 into active caspase 1 and the downstream production of IL-1β and IL-18. Pyrin, the protein defective in familial Mediterranean fever is a negative regulator of this process. Defects in NALP3, also know as CIAS1, are associated with NOMID, Muckle-Wells syndrome, and familial cold urticaria. The mutations in these cases are activating mutations that drive chronic production of IL-1β and IL-18. As may be seen, this pathway also affects apoptosis.

T Cells must also undergo an education process to prevent the development of autoreactive cells. T-cell progenitors enter the thymus and can develop into T cells bearing either the γδ or the αβ T-cell receptor. Much less is known about γδ T-cell development than αβ T-cell development. The αβ T cells develop in the thymus, responding to chemokine signals. They proliferate vigorously in the subcapsular zone as double-negative T cells, because they lack both CD4 and CD8.22 At this point, the T cells undergo rearrangement of their T-cell receptor genes (eFig. 198.1 ). There is an interaction of T cells and thymic stromal cells that is required to develop the corticomedullary structure, suggesting that signals travel both from the thymic epithelium and to the thymic epithelium. Once a successful rearrangement has occurred, the T cells begin to express both CD4 and CD8 and these T cells are termed double-positive (DP) T cells. It is these cells that undergo the rigorous selection process. Initially the highly motile DP T cells interact with stromal cells.23 T-cell receptors capable of interacting with self-MHC receive a survival signal, whereas T cells with no ability to recognize self die of neglect. The DP T cells that survive this first screening migrate to the medulla, where they test their receptor against a variety of tissue-specific antigens. These are induced by at least one transcription factor designed to upregulate genes that would not otherwise be expressed in the thymus. The thymus becomes a “mirror of self” and thus negative selection of self-reactive T cells can occur. Regulatory T cells also develop in the medulla. Further development in the medulla also leads to the expression of either CD4 or CD8, in such a way that the cells become single positive T cells. The single positive T cells exit the thymus into the periphery, where they recirculate via lymphatics.24 T cells only recognize antigen in the context of MHC. CD4 T cells recognize antigen in the context of MHC class II; CD8 T cells recognize antigen in the context of MHC class I. The antigen-presenting cell is more than a passive repository of antigen; it must also deliver a second signal to instruct the T cell as to whether it should treat the antigen as a threat.25-27 Dendritic cells, macrophages, and B cells are the main antigen-presenting cells. Each of these can upregulate costimulatory molecules, which provide the second signal for T cells. Often the signal for upregulation of the costimulatory molecules is the pathogen itself, recognized by a TLR. Alternatively, inflammatory cytokines can also induce costimulatory molecules. Antigen plus a costimulatory signal drive the T cell to proliferate and ultimately to execute its effector function, including production of cytokines, providing help for B cells, and killing infected targets. In addition, T cells provide critically important regulatory functions for other T cells and other immunologically competent cells. Antigen in the absence of a costimulatory signal drives the T cell to become anergic and unable to proliferate.28 In the absence of a perceived threat, the T cell presumes the antigen is self and ignores it. In infections in which the danger signal is not spatially limited, bystander T cells could become inadvertently activated, potentially leading to a break in tolerance.

MECHANISMS OF TISSUE DAMAGE

Pathologic host immune responses are often divided into four categories, although it is important to remember that these processes are part of a complex pathogen defense strategy as well. Immediate or type I hypersensitivity refers to the typical allergic response. Preexisting IgE directed against an antigen is bound to mast cells. Engagement of IgE by the antigen leads to immediate degranulation of mast cells with release of histamine as well as a variety of cytokines. IgE-mediated tissue damage is seldom relevant in autoimmune conditions, but there are limited data suggesting that mast cells can modulate inflammation.

Type II hypersensitivity refers specifically to autoantibody-mediated tissue damage. The membrane attack complex of the complement cascade primarily harms cells such as erythrocytes, which have no ability to repair membrane damage. Autoantibodies can also lead to tissue damage by facilitating uptake by the reticuloendothelial system. An example of this is idiopathic thrombocytopenia purpura. In other cases, the autoantibodies serve as an opsonin for the self-tissue and lead to the recruitment of neutrophils and macrophages, which then induce tissue damage. This type of autoantibody-mediated damage occurs in rheumatic fever, where antibodies to Streptococcus cross-react with myocardial antigens, leading to inflammation and symptoms.

Finally, autoantibodies can induce harm by acting as antagonists for a receptor or in some way interfering with receptor function, as is seen in myasthenia gravis. Immune complex disease or type III hypersensitivity is best exemplified by serum sickness. Originally described in the course of passive immunization to diphtheria, it now most often occurs secondary to drug use. Classically, approximately 7 to 10 days after beginning a medication, as antibody production begins, the patient develops fever, arthritis, and proteinuria. A vasculitic rash sometimes also may be seen. Skin, joint, and kidney are characterized by small arterial beds with high oncotic pressure, the anatomic region most likely to be involved during serum sickness. This high oncotic pressure leads to deposition of immune complexes in involved areas, in turn activating Fc receptors locally. This then leads to the production of inflammatory cytokines and recruitment of neutrophils and other inflammatory cells. Some manifestations of systemic lupus erythematosus may be mediated through this process. Cryoglobulinemia also leads to an immune complex–like process.

Delayed or type IV hypersensitivity is familiar to clinicians as the basis of the PPD test for tuberculosis. Here, T cells are responsible for the tissue damage. Although a break in T-cell tolerance is hypothesized to underlie many autoimmune diseases, there are few disorders in which the main mechanism of tissue damage is believed to be T-cell driven. Despite the association of many autoimmune diseases with MHC haplotypes, dramatic infiltrates of T cells are seen in relatively few diseases of childhood. Examples include multiple sclerosis, diabetes mellitus, and Crohn disease. In juvenile idiopathic arthritis, the synovial fluid often contains high numbers of neutrophils, a testament to the role of the innate immune system in the signs and symptoms of the disease. In the synovial membrane, the infiltrating cells are largely macrophages and T cells. In some cases, infiltrating cells adopt a lymph node–like architecture, including a germinal center. This lymphoid aggregate is believed to require a T-cell contribution, but the exact role of the T cell in this complex process is not known. One theory is that the recognition of antigen by T cells may drive cytokine production, thereby leading to many of the downstream effects.

INFLAMMATION

MICROCIRCULATION CHANGES

Clinicians identify inflammation on the basis of very obvious manifestations: erythema, tenderness, induration, warmth, and often loss of function. These features arise owing to changes in microcirculation.29In the setting of autoimmune or autoinflammatory disorders, effects of inflammation seem largely pathologic. When viewed as a response to infection, however, these changes are more physiologically comprehensible.

Regardless of the cause of inflammation, damaged capillaries activate Hageman factor, also known as factor XII, of the coagulation pathway. Hageman factor activates factor XI and initiates coagulation at the site of tissue damage. This process stems bleeding but also traps platelets and slows the flow of blood to the site. This, in turn, facilitates neutrophil adhesion. Hageman factor also activates the kallikrein pathway leading to bradykinin production. Bradykinin is responsible for the vascular leak that leads to induration and swelling. It is also one of the mediators of vascular dilatation, causing the characteristic redness seen in inflammation. The kallikrein pathway can also directly cleave complement protein C5, a potent chemotactic factor that attracts neutrophils to the site.30 These microcirculatory changes provide the optimal setting for the recruitment of neutrophils.

Bacteria interact with TLRs on the surface of sentinel cells, leading to the production of TNF-α. This cytokine increases P-selectin and E-selectin expression by the vascular endothelium. These selectins bind P-selectin glycoprotein ligand-1 (PSGL1) and other glycoproteins to initiate the rolling phase of neutrophil adhesion. Inflammatory cytokines, platelet-activating factor, and chemokines can activate the β2 integrins, which mediate the firm adhesion of the neutrophil to the vascular endothelium. This process is aided by the slow flow of blood in the inflamed site. The arrested neutrophil can undergo diapedesis across the vascular endothelium and chemotax toward the bacteria. Vascular leak allows complement and antibody to escape from the blood and join neutrophils in the tissue, facilitating opsonization and ingestion of bacteria. Thus, each of the pathologic changes in inflammation has a clear physiologic benefit.

TOLERANCE

Tolerance is the word for the process whereby T cells and B cells are instructed to avoid self-responses. The processes for both T and B cell tolerance are incompletely understood; however, recent insights from single-gene disorders have led to a fuller picture of the delicate nature of establishing tolerance.

B-CELL DEVELOPMENT AND TOLERANCE

B cells develop in the bone marrow and the B-cell receptor, or surface immunoglobulin, is generated via a series of recombination events. The events lead to random generation of receptors, only a few of which will be useful in host defense. The heavy chain initially undergoes rearrangement at the pro–B cell stage. This heavy chain is expressed with surrogate light chains and at this stage the ability of the heavy chain to pair appropriately with a light chain is tested. An effective pairing stimulates light chain rearrangement, which is completed at the pre–B cell stage in the bone marrow. A functional B-cell receptor exists at this phase of B-cell development, and it becomes fully expressed on the cell surface at the immature B-cell stage. The B-cell receptor is expressed with two proteins, Igα and Igβ, which transmit signals to the interior of the cell. At the immature B-cell stage, strong engagement of the B-cell receptor leads to apoptosis. Complement proteins appear to play a role in this important process because complement deficient mice have impaired B-cell tolerance and complement deficient humans (C1, C2, C4) are more prone to development of lupus.

This process serves to prevent autoreactive B cells from reaching the periphery, but it is not particularly efficient, perhaps because relatively few self-antigens are expressed in the bone marrow environment. Another option for deleting autoreactive B cells, known as receptor editing, also takes place at the pre–B to immature B-cell stage. Approximately 25% of peripheral B cells appear to have undergone receptor editing, which consists of an additional round of light chain recombination that eliminates autoreactive B-cell receptors.37 Studies have shown that 55% to 75% of all early immature B cells express autoreactive receptors. The processes of clonal deletion and receptor editing reduce the frequency to approximately 40%, meaning that a large number of autoreactive cells still manage to leave the bone marrow.16

There is a further checkpoint in B-cell development. The number of autoreactive mature B cells is further diminished to 20% and this is believed to be the result of competition for niches in the secondary lymphoid organ follicles. Even though these checkpoints eliminate 80% of autoreactive B cells, a relatively large number of potentially pathologic B cells remain. In addition, patients with rheumatoid arthritis and lupus have been shown to have defective checkpoints such that even larger numbers of cells escape deletion.38,39

In general, the autoreactive B cells do not cause harm as long as they do not receive T-cell help. This process can be short-circuited by infection, explaining why autoimmune diseases may be precipitated or exacerbated by infections. Marginal zone B cells appear to be particularly prone to activation by bacterial products via TLRs. Conversely, viral infections may activate autoreactive B cells via the production of type I interferons, which stimulate B-cell proliferation and antibody production.40 The large number of circulating autoreactive B cells and the high frequency of infections suggest that additional, thus far unidentified, checkpoints must exist to prevent autoreactive B cells from producing autoantibody.

T CELL DEVELOPMENT AND TOLERANCE

As described in Chapter 186, T cells develop in the thymus (eFig. 198.1 ). Similar to B cells, they initially generate a heavy chain, which is expressed with a surrogate alpha chain protein. This occurs in the double-negative (DN) T-cell stage. The alpha chain subsequently undergoes rearrangement and the cells progress to become double-positive (DP) T cells. Positive selection for cells that have some capacity for self-recognition occurs in the cortex. This initial screening process ensures that the receptor will engage self-MHC. Unlike B cells, T cells continue to undergo rearrangements until they receive a positive signal. Recombination here too leads to random receptor specificities and the majority of T cells fail to produce a T-cell receptor that can positively interact with self-MHC. Consequently, the majority of T cells die at this stage.

Medullary epithelial cells express organ specific antigens. A transcription factor termed AIRE induces the expression of approximately 500 to 1200 genes in the thymus.41 Each of these genes encodes proteins whose expression would otherwise be limited to a distant tissue. In this way, the double-positive T cell can survey a set of antigens it is likely to encounter in the periphery. Autoreactive T cells at this point undergo apoptosis. Defects in the AIRE gene are associated with the progressive accumulation of autoimmune processes, a disorder called autoimmune polyendocrinopathy, candidiasis, ectodermal dysplasia (APECED).42 This condition vividly demonstrates the importance of negative selection in the thymus.

Murine models, in which manipulation is easier, demonstrate that the contribution of antigen to development of T cells is also critically important. Diminished expression of insulin in the thymus leads to an increased frequency of diabetes in a murine model, probably because the T cells cannot be adequately tolerized.43 These antigens are all presented to T cells in the context of an MHC molecule. An unstable MHC provides a poor platform for antigen presentation; markedly diminished expression of MHC can lead to both poor positive selection and poor negative selection.44 Thus, patients with an inherited immune deficiency that causes diminished MHC class I expression (the bare lymphocyte syndrome) typically have very few CD8 T cells, because CD8 T cells are positively selected though MHC class I.45These patients also have a high frequency of autoimmune diseases. Similarly, patients with markedly diminished MHC class II expression have few CD4 T cells and also have a relatively high frequency of autoimmune disease. As one would expect, stem cell transplantation is not particularly effective in these conditions unless the graft can provide some antigen presenting cells to populate the thymus.

Another type of αβ T cell is the regulatory T cell.46 These cells have particularly high affinity for self and enter a separate developmental pathway dependent on the transcription factor FoxP3. These cells were originally identified in neonatally thymectomized mice. The mice had small numbers of conventional T cells, but they all died of severe autoimmune disease characterized by massive T-cell organ infiltration. Replacement of CD25-positive T cells prevented development of the autoimmunity. Rare patients who have mutations in the critical transcription factor FoxP3 demonstrate the importance of this pathway in humans.47-50 They develop immune deficiency, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX), which is manifested by a T-cell infiltrate in the small bowel and accumulation of other autoimmune organ involvement. The process is typically fatal in the absence of a stem cell transplant. The role and potential manipulation of regulatory T cells in human diseases is now an active area of investigation.47-49

The balance of antigen presentation, thymic selection and the development of cells destined to provide a safety net in the periphery (regulatory T cells) interact to prevent autoimmune disease. As is the case for B cells, environmental factors appear to affect the process, although these are not well understood.

GENETIC FACTORS CONTRIBUTING TO AUTOIMMUNITY

A break in tolerance is believed to initiate synovitis, although the ultimate joint damage is owing to the collaborative effects of many downstream pathways (Fig. 198-2). The cellular interactions defined in joint inflammation may be partially extrapolated to other types of autoimmunity. Still there are significant differences between various types of autoimmune pathology, and much to learn about the subtle differences that can mold a particular disease phenotype.

PATHOGENIC MECHANISMS IN COMPLEX POLYGENIC DISORDERS

Juvenile idiopathic arthritis (JIA) is the most common pediatric rheumatologic disorder. In spite of this, little is understood concerning its genetics. The relative risk in siblings, a rough estimate of the genetic contribution to the disease, is approximately 15, similar to that of multiple sclerosis and insulin dependent diabetes mellitus.51 The concordance among twins is approximately 25%, which also suggests a strong genetic component.52 These epidemiologic analyses have also identified a strong association of JIA with diabetes, suggesting that some genetic factors may lead to a more global predisposition to autoimmunity.51 Unfortunately, epidemiologic studies have generally grouped all forms of JIA together, so it is difficult to know whether the genetic contribution to oligoarthritis differs significantly from that of the polyarticular form. In addition, there are significant racial and ethnic differences in the incidence of arthritis. These could relate to either environmental or genetic contributions,53-56 including the HLA variations between different human populations (eFig. 198.2 ).

FIGURE 198-2. Major inflammatory pathways in the joint. Bone destruction, seen in a minority of patients with JIA, is mediated predominantly by osteoclasts. These cells are related to macrophages but are part of normal bone homeostasis. With increased IL-17, MCSF, and RANKL, the osteoclasts become disproportionately activated compared with bone development and erosions occur. The inciting antigen is not known, but IL-17-producing T cells (Th17 cell) migrate into the joint and are present in the synovium. There is a complex interplay of the Th17 cells and synovial macrophages that is contact dependent and leads to activation of macrophage expression of TNF-α, IL-6, and IL-1β. These cytokines contribute to osteoclast activation but also very importantly stimulate synovial fibroblast proliferation and activation. These cells form the structure of the synovial hypertrophy, often seen in JIA. These activated synovial cells produce MCSF and RANKL, which contribute to osteoclast activation, and TNF-α and IL-1β, which further activate the synovial macrophages. Finally, angiogenesis, which is required to support the metabolism associated with growth and activation, is aided by activated macrophage production of vascular endothelial growth factor (VEGF). IL, interleukin; TNF, tumor necrosis factor.

The epidemiologic studies in JIA certainly suggest a genetic component, and the same is true for pediatric lupus and dermatomyositis. In the case of lupus, the sibling recurrence risk is 20 and the twin concordance is 24% to 65%.57,58 In the case of pediatric dermatomyositis, no sibling studies have been done, but there is a strong MHC association.

The association of various pediatric rheumatologic conditions with MHC polymorphisms has two important implications. First, antigen is presented to T cells in the context of MHC, suggesting that autoreactive T cells are driving much of the pathologic autoimmunity. Supporting this idea is the fact that T cells are found in the synovium of JIA patients and in the muscle of patients with dermatomyositis. Nonetheless, T-cell-targeted interventions have not been widely utilized, and when tried, have not been particularly effective for treating pediatric rheumatologic conditions. This could be owing to the very strong linkage disequilibrium of the MCH region; the association of disease with an MHC type may in fact be the result of linkage with another gene in the region.

The second implication of the finding that most pediatric rheumatologic conditions are associated with MHC genes relates to the fact that there are multiple genes that contribute to overall genetic risk. In most cases, the relative risk associated with inheritance of any individual MHC allele is approximately 2 to 3; other genes thus must contribute to the overall genetic risk. Perhaps best studied in murine models of lupus, individual loci appear to contribute one of a series of steps that together lead to the development of lupus.59 Inheritance of one locus confers increased B-cell activation and inheritance of another leads to increased antigen presentation. The implication is that individual risk factors are cumulative.60

An aspect related to genetics, but which is not inherited, is called epigenetics. This refers to features of the genome that are not encoded by the DNA. Typically this is taken to mean the packaging of DNA into accessible or inaccessible regions. The sequence of the DNA may be the same in twins, but the expression of individual genes may differ significantly between the two because of epigenetic factors.61 This is not a simple issue of semantics; deliberate alteration of epigenetic characteristics has led to the development of lupus in two distinct models.62,63

EVIDENCE SUPPORTING AN ENVIRONMENTAL CONTRIBUTION

Potential environmental factors involved in the etiopathogenesis of JIA include high income, no siblings, and urban living.64 In general, JIA does not have seasonal clustering, although in some studies systemic JIA was found to occur more often during colder months.65 The most strongly seasonal pediatric rheumatologic disease is Kawasaki syndrome, in which many studies in different geographic regions demonstrate a strong winter–spring seasonality.66There have also been studies demonstrating temporal clustering, a finding that is also consistent with an infectious etiology. Although many studies have attempted to identify an infectious origin of this condition, results to date have been inconclusive.

One of the more intriguing findings in lupus is the fact that pediatric patients are more likely to have been infected with Epstein-Barr virus (EBV) than were controls.66 Coupled with strong data demonstrating that type I interferons drive B-cell development, and the recognition that type I interferons are produced preferentially after viral infections, this may provide an explanation for flares after common infections, and perhaps an explanation for the initial insult itself. On the other hand, EBV infection is ubiquitous in virtually all populations, yet lupus develops in fewer than 10 children in 100,000. Further, lupus tends to evolve over a 5- to 10-year period. Thus, whereas infection may contribute to disease initiation, there must be other mechanisms that move the process forward. Estrogen, medications, and sun exposure have been implicated as environmental cofactors, but the process of disease evolution is not well understood.

CONTRIBUTIONS OF THE INNATE IMMUNE SYSTEM TO PEDIATRIC RHEUMATOLOGIC CONDITIONS

One explanation for the development of most types of disease, including autoimmunity, is to hypothesize that patients inherit a genetic predisposition. This predisposition, whether strong or weak, interacts with environmental and stochastic events and eventually leads to pathology. In the case of autoimmune diseases, the intermediate event is amplification of an autoreactive B-cell or T-cell clone. This clone, in turn, is regulated by a more-or-less robust homeostatic process than may or may not keep it under control. If control is breached, the clone continues to expand and disease ensues. The disease manifests itself after recruitment of varied inflammatory cells. Some autoimmune phenomena are naturally self-limited, suggesting that a mistake need not become a chronic disease. Examples of short-lived autoimmune phenomena include idiopathic thrombocytopenia purpura, Henoch Schonlein purpura, some vasculitides, and autoimmune hemolytic anemia. In many cases, however, the process does become chronic, and the processes that perpetuate the disease need not be the same as those that precipitate it. The end organ infiltrate may be composed of mixed lymphocytes and phagocytic cells that were recruited long after the inciting event.67