Immunopathogenic Background of Pars Planitis

Pars planitis is defined as an intermediate uveitis of unknown background of systemic disease with characteristic formations such as vitreous snowballs, snowbanks and changes in peripheral retina. The incidence of pars planitis varies 2.4–15.4 % of the uveitis patients. The pathogenesis of the disease is to be determined in future. Clinical and histopathological findings suggest an autoimmune etiology, most likely as a reaction to endogenous antigen of unknown source, with T cells predominant in both vitreous and pars plana infiltrations. T cells subsets play an important role as a memory-effector peripheral cell. Snowbanks are formed as an effect of post inflammatory glial proliferation of fibrous astrocytes. There is also a genetic predisposition for pars planitis by human leukocyte antigen and several other genes. A coexistence of multiple sclerosis and optic neuritis has been described in numerous studies. Epiretinal membrane, cataract, cystoid macular edema, retinal detachment, retinal vasculitis, neovascularization, vitreous peripheral traction, peripheral hole formation, vitreous hemorrhage, disc edema are common complications observed in pars planitis. There is a need to expand the knowledge of the pathogenic and immunologic background of the pars planitis to create an accurate pharmacological treatment.


Introduction
Intermediate uveitis described by the Standardization of Uveitis Nomenclature is the inflammation in the anterior vitreous, ciliary body and the peripheral retina, the vitreous is the major site of the inflammation. If there is a systemic disease (e.g., sarcoidosis) or associated infection (e.g., Lyme disease) (Bloch-Michel and Nussenblatt 1987;Jabs et al. 2005), the term idiopathic should not be used. There are many names that have been used to describe inflammation in the anterior vitreous, ciliary body and peripheral retina such as intermediate uveitis, pars planitis, chronic cyclitis, peripheral uveitis, vitritis, cyclochorioretinitis, chronic posterior cyclitis and peripheral uveoretinitis (Foster and Vitale 2013).
The diagnostic term ''pars planitis'' should be used only for an intermediate uveitis of unknown systemic disease (absence of any known systemic cause) or not related to any infection and where the snowbank or snowball formation is observed (Jabs et al. 2005).
There are limited data on incidence of pars planitis itself. Usually the reported epidemiology underlines incidence of intermediate uveitis (varying between 1.4 and 22 %) as a wider group and not pars planitis (Chan et al. 2007;Foster and Vitale 2013;Henderly et al. 1987;McCannel et al. 1996;Rodriguez et al. 1996).
The pathogenesis of the disease is to be determined in future. Clinical and histopathological findings suggest an autoimmune etiology, most likely as a reaction to endogenous antigen of unknown source, with T cells predominant in both vitreous and pars plana infiltrations. There is limited number of research where the study group includes only patients with pars planitis. Usually research includes patients with intermediate uveitis to the study group and sometimes subtracts a subgroup of patients with pars planitis.
The course of pars planitis and possible complications are well defined and were observed in many trials (Babu and Rathinam 2010;Donaldson et al. 2007;Eichenbaum et al. 1988;Felder and Brockhurst 1982;Foster and Vitale 2013;Green et al. 1981;Jabs et al. 2005;Kenyon et al. 1975;Malinowski et al. 1993a;Nussenblatt and Palestine 1989;Pederson et al. 1978;Pruett et al. 1974;Smith et al. 1976). However, the pathophysiology and possible impact of immune system to the course of the disease and to the complication are to be determined.
Numerous papers were found but only few matched both criteria (approximately 25 items were founded). During the search for ''experimental models of uveitis'', task was easier to accomplish and items that were founded were numerous; approximately 17 of them were accurate for probable experimental models of intermediate uveitis or pars planitis. Founded papers focused on different aspects of the disease and define the study group differently. Most of them as patients with intermediate uveitis, some are subtracting a subgroup including patients suffering from pars planitis and few define study group as patients with pars planitis in the strict sense of this term presented previously.
Obtained data will be presented in sections entitled: epidemiology, pathogenesis, clinical features, complications and summary.

Epidemiology
Intermediate uveitis has been reported in 1.4-22 % of uveitis patients and there are limited data on the incidence of pars planitis (Chan et al. 2007;Foster and Vitale 2013;Henderly et al. 1987;McCannel et al. 1996;Rodriguez et al. 1996). The rate of 1.5 per 100,000 per year was reported by The Northern California Epidemiology of Uveitis Study (Gritz and Wong 2004). In tertiary referral centers, 2.4-15.4 % of uveitis patients have been reported as the one with pars planitis diagnosis, but it must be taken under consideration that these centers observed more severe end of the spectrum of ocular inflammatory disease (Henderly et al. 1987;Wakefield et al. 1986).
Histopathological and clinical findings suggest a possible autoimmune etiology of pars planitis. In the infiltrate surrounding of the retinal vessels and in the snowbanks, the Th cells have been found. They can be possibly directed against some unknown ocular antigens (Wetzig et al. 1988;Yoser et al. 1992). The vitreoretinal interface, associated with retinal blood vessels, ciliary body, or the pars plana itself, may be the source of sought antigens (Davis et al. 1992).
Several experimental models of uveitis were described, used especially in cases of posterior uveitis, induced by a subcutaneous injection of single dose of soluble antigens (usually retinal). Mentioned models (see Table 1) include: experimental autoimmune uveitis (EAU) (Adamus and Chan 2002;Chan et al. 1985a, b;de Kozak et al. 1981), experimental melanin protein-induced uveitis (EMIU) (Adamus and Chan 2002;Matteson et al. 1999), experimental autoimmune encephalitis associated to anterior uveitis (EAE/AU) (Adamus et al. 2000;Adamus and Chan 2002;Buenafe et al. 1998) and endotoxin-induced uveitis (EIU) (Li et al. 1995). In each model, different soluble antigen is used; S-arrestin or interphotoreceptor retinoidbinding protein (IRBP), melanin proteins, myelin basic protein and lipopolysaccharide, respectively, for EAU, EMIU, EAE/AU and EIU (Gasparin et al. 2012). None of the described models fully explains the changes that occur in humans, especially in case of pars planitis. The possible candidates for pars planitis antigens were specified in conducted experiments involving humans. Those include human lens protein fraction (Doycheva et al. 2007), HNK-1 carbohydrate epitope (Uusitalo et al. 2000) and beta B1crystallin (Stempel et al. 2003).
The exact cause of pars planitis remains unknown (see Table 2), but CD4 ? T cells that express an activation marker CD69 (an early activation marker) in either peripheral blood (Murphy et al. 2004) or in the aqueous humor are found, respectively, in patients with pars planitis and with idiopathic uveitis (Calder et al. 1999). Longitudinal measurements over a period of treatment rather than single evaluation of CD4 ? CD69 ? T cells may be a useful substitute marker of ocular inflammation. As such it may be used to asses disease activity, but further studies are needed (Kilmartin et al. 2001;Murphy et al. 2004).
There are some studies showing increase in levels of surrogate markers of immune activation in the serum of patients with intermediate uveitis. These markers include IL-8 and soluble intracellular adhesion molecule 1 (Klok et al. 1998(Klok et al. , 1999 as well as increased expression on peripheral CD4 ? T cells such as soluble IL-2 receptor, HLA-DR and CD25 of patients with uveitis (Deschênes et al. 1988;Dick et al. 1992;Feron et al. 1995). All of mentioned above need to be confirmed in longitudinal studies before they become useful in clinical practice, in predicting an outcome of the disease and monitoring of the immunosuppressive treatment.
Studies of the model of experimental autoimmune uveitis have shown that Th1-related cytokines predominate in the active phase of the disease and IL-10 during resolution phase, indicating a shift towards Th2 (Barton et al. 1995). Further studies of peripheral blood CD4 ? T cells expression of chemokines show no difference between the Th1-associated chemokine receptors CXCR3 and CCR5 or the Th2-associated chemokine receptor CCR4 in patients suffering from intermediate uveitis (enrollment of patients includes idiopathic and presumed sarcoid intermediate uveitis) compared to healthy controls as well as active compared to inactive disease (Murphy et al. 2004). Murphy et al. (2004) described an increase in the intracellular CD4 ? T cell expression of TNF-a in idiopathic intermediate uveitis which suggests a predominant CD4 ? T cell subset and a new potential role of anti-TNF-a treatment in this type of uveitis (Sugita et al. 2006a).
In animal models of experimental autoimmune uveoretinitis, a transfer of CD4 ? CD25 high Foxp3 ? Treg cells confer protection from uveitis induced by the uveitogenic retinal antigen IRBP (Keino et al. 2007;Siepmann et al. 2007;Silver et al. 2007). The lower percentage of CD4 ?-Foxp3 ? T cells was observed in patients with active uveitis compared to inactive. However, no statistically significant differences were found in percentage of CD4 ? Foxp3 ? T cells between patients with known systemic autoimmune disease compared to idiopathic uveitis. The outcome was this same for patients with cystoid macular edema compared with no evidence of cystoid macular edema. In addition, no linear correlation of the percentage of CD4 ?-Foxp3 ? T cells with logMAR best corrected visual acuity was observed (Yeh et al. 2009).
Until now, little is known about the role of CD8 ? T cells in pars planitis, and of coexpression of functional NK receptors, like CD56 or CD57 on CD8 ? T cells, which are thought to be associated with effector function of these cells (Ahn et al. 2005;Tarazona et al. 2000). Particularly CD57 expressed on both CD4 ? and CD8 ? T cells is considered to be correlated with late immune responses (d'Angeac et al. 1994). Many studies have described an association between CD57 expression and persistent antigenic stimulation and chronic diseases. Pedroza-Seres et al. (2007) described in their paper a subset of CD57 ? T cells. CD57 ? T cells had a peripheral memory phenotype (CCR7 -CD27 -CD28 -). The role of these cells in helper cell regulation was evaluated by cytokine production in CD57 ? T cells. The intracellular proteins involved in cytotoxicity were measured in CD8 ? CD57 ? T cells and its elevated expression may suggest that these cells play an effector role in pars planitis (Pedroza-Seres et al. 2007). Also, in the model of experimental autoimmune uveitis, CD8 ? T cells can be stimulated by IRPB and TGF-b1 to express Foxp3 in high levels and may have suppressive functions (Peng et al. 2007).
Role of B cells was evaluated in studies performed on enucleated eyes, in case of infiltration of these cells in pars plana, vitreous and retina, but the results were not compared to immunological findings in the peripheral blood. In Later stages the ratio changes to 1/1, or even IL-2, IFN-c and TNF-a are produced by the subset Th1 Adamus and Chan (2002); Chan et al. (1985a, b); de Kozak et al. (1981) Experimental melanin protein-induced uveitis

Melanin protein
Uveal melanocytes CD4 ? T cells and macrophagesthe primary mediators of the inflammatory response Early stages-IL-2 and IFN-c Medium stage-IL-12 The peak of the inflammation-TNF-a Other inflammatory mediators such as nitric oxide can be also involved Adamus and Chan (2002); Matteson et al. (1999) Experimental  addition, the enucleated eyes were in last stage of the disease and nowadays with more efficient immunosuppressive treatment and wider range of possible surgical options, the stages usually are not recorded. Finally, the researches were conducted in the eighties of twentieth century and diagnostic possibilities have now improved. The aspect of innate response in pars planitis was not taken into account in conducted researches, so far. The role of macrophages and neutrophils has not been evaluated so far. In the light of newly described innate lymphoid cells (ILCs) that have a function analogic to Th cells while also including NK cells, there will be a great demand of further research on ILCs in pars planitis.
From the point of view presented above, better understanding of immunological processes taking place in an affected eye during pars planitis may play a key role in pathogenesis of pars planitis and in appropriate treatment of this disease.

Genetic Background
There is also a genetic predisposition for the pars planitis. Number of studies described the contribution of the human leukocyte antigen (HLA) and several other genes in addition to those of the HLA in the pathogenesis of pars planitis. However, they differ in inclusion and exclusion criteria used to eliminate patients with intermediate uveitis due to other causes (e.g., sarcoidosis, syphilis) and described differences may play an important role why results of studies vary from each other. Several studies claimed the contribution of the HLA alleles considered as susceptibility markers in multiple sclerosis and optic neuritis which are: HLA-DR15, -DR51 and -DQ6 (Hillert et al. 1996;Malinowski et al. 1993b;Oruc et al. 2001;Raja et al. 1999;Tang et al. 1997). The subtype of intermediate uveitis was more often found in patients with multiple sclerosis (Bamford et al. 1978;Giles 1970;Porter 1972). An elevated frequency of HLA-B8, -B51 and -DR-2 was estimated after exclusion of patients with multiple sclerosis from the pars planitis group (Malinowski et al. 1993b). Some authors claimed that HLA-DR15, -DR51 and -DR17 contributed to the etiology of both multiple sclerosis and pars planitis (Oruc et al. 2001;Raja et al. 1999). Some studies evaluated a genetic predisposition of pars planitis and its course, for example, in Mexicans. HLA-A, -B, and -C typing was done on T cells isolated with immunomagnetic beds from a group of untreated pars planitis patients with Mexican origins. The significant increase in frequency was found in DRB1*0802 in these patients. There was an association between clinical features, gender and HLA alleles (shown in Table 3; Alaez et al. 2003). Nonetheless, there can be found some studies that show no correlation between pars planitis and HLA alleles (Greiner et al. 2003;Mantrana-Bermejo et al. 2010).

Clinical Features
Patients with pars planitis usually present with minimal symptoms such as floaters or blurred vision. However, in severe cases, a significant visual loss can occur due to aggregation of floaters in the vitreous or due to macular oedema. The signs of inflammation from the anterior chamber may be minimal in the form of keratic precipitates, flare and cells or they may be absent. Sometimes we can also find posterior synechiae which are usually seen involving the inferior iris. A characteristic feature of pars planitis-vitritis is typically described and evaluated by clinicians as vitreous haze ranging from trace to 4? (Jabs et al. 2005).
The vitreous snowballs are yellow-white inflammatory aggregates which can be found in the midvitreous and inferior periphery (Felder and Brockhurst 1982;Nussenblatt and Palestine 1989;Pruett et al. 1974) (Figs. 1, 2). Some studies described vitreous snowballs as made of granulomatous inflammation (Green et al. 1981). The difficulty to get the material for further research precisely from intended spot results in lack of many molecular studies.
Snowbanks, the snow-white exudates that are found on the pars plana, usually are seen in the inferior area but they also can extend to 360°of the retinal periphery. Typically they are associated with more severe forms of the disease, and require more aggressive therapy (Felder and Brockhurst 1982;Nussenblatt and Palestine 1989;Pruett et al. 1974). As previously described in histopathologic studies, snowbanks consist mainly of glial elements (probably fibrous astrocytes which can produce a large diameter collagen fibrils and basement membranes) (Eichenbaum et al. 1988;Green et al. 1981;Kenyon et al. 1975;Pederson et al. 1978), it was confirmed by staining with anti-Muller cell antibody and anti-glial fibrillary acid protein (GFAP). The formation of snowbanks is stimulated by inflammation. Activated Muller cells express class II MHC molecules and present antigen to T cells, sustaining an active state of inflammation itself (Roberge et al. 1985). The major collagen glycoproteins of snowbanks are type IV collagen and laminin (formed rather by glial cells than fibroblasts as the letter one's characteristic products are collagen type I and III). All these facts confirmed a hypothesis of postinflammatory glial proliferation that takes place in eye of pars planitis patients. Different results from the one's described above can be found in Abu El-Asrar and Geboes (2002) work. In their point of view, the snowbanks are an acellular formation (in exception of Table 3 Distribution of HLA alleles in pars planitis with association to clinical features (family reports were excluded)  Retinal changes in pars planitis occur in its peripheral part and include tortuosity of arterioles and venules, sheathing of veins (Fig. 3), neovascularization and sometimes retinal detachments (Felder and Brockhurst 1982;Nussenblatt and Palestine 1989;Pruett et al. 1974). The vascular sheathing in histopathologic studies presents a cellular infiltration dominated by T cells with few B cells usually posterior to the junction of the iris and ciliary body (Eichenbaum et al. 1988;Wetzig et al. 1988).

Complications
The most common ocular complication in patients with pars planitis is epiretinal membrane (ERM) with high incidence from 7 to 36 % in different studies. The mean interval between onset of pars planitis and ERM was 6.5-7.9 years (Babu and Rathinam 2010;Donaldson et al. 2007;Malinowski et al. 1993a). The mechanism of epiretinal membrane formation during pars planitis is unknown.
Cataract was also common for pars planitis (frequency 14-30 %), it occured as a result of the disease and the treatment. It was difficult to distinguish between these too situations retrospectively. Most of the diagnosed cataracts were posterior subcapsular cataract followed closely by mixed nuclear sclerotic and posterior subcapsular cataracts. Described complications were developed much later, mean onset after presentation was 9.8 years (Babu and Rathinam 2010;Donaldson et al. 2007;Foster and Vitale 2013;Malinowski et al. 1993a;Pruett et al. 1974;Smith et al. 1976). The mechanism of steroid cataract induction remains unknown, nevertheless the observation of glucocorticoid-induced gene transcription and the finding of a transcriptionally active glucocorticoid receptor in human lens epithelial cells suggest that glucocorticoids can act directly in lens epithelial cells and modulate several signaling pathways influencing gene transcription (Dickerson et al. 1997;Gupta and Wagner 2009;James et al. 2003).
Cystoid macular edema (CME) was described as a major cause of visual loss (incidence 8-26 %) in patients with pars planitis (Fig. 4). The mean interval between onset of pars planitis and CME was 5.7 years (Babu and Rathinam 2010;Donaldson et al. 2007;Foster and Vitale 2013;Malinowski et al. 1993a;Pruett et al. 1974;Smith et al. 1976). The pathophysiology of cystoid macular edema was Fig. 1 16-year-old feme patient with diagnosed pars planitis, first ophthalmological examination. Snowballs are seen as focal yellowwhite massive exudates in peripheral anterior vitreous  Fig. 3 17-year-old female patient with diagnosed multiple sclerosis, first ophthalmological examination. Vascular sheaths also known as vascular cuffs are seen as massive perivascular exudates described in patients with uveitis without subdividing that group (no data show a pathomechanism of CME in patients with pars planitis). Retinal vascular hyperpermeability can be influenced by several mediators such as: nitric oxide, TNF-a, vascular endothelial growth factors, prostaglandins, leukotrienes, cytokines IP-10 and IL-6 (Acharya et al. 2009;Curnow et al. 2005;Markomichelakis et al. 2005;Probst et al. 2004;van Kooij et al. 2006). Vascular leakage could result from endothelial cell damage (Leal et al. 2007).

Conclusion
Little is known about background and pathology of pars planitis. From the patient and the ophthalmologist point of view, there is a great need for immunological research of this condition to improve the treatment.