Chronic Intraocular Leptospiral Infection Relying on Biofilm Formation inside the Vitreous Cavity Leads to Recurrent Uveitis in Horses

Abstract

Equine recurrent uveitis (ERU) is a disease known and feared for centuries, as it almost always leads to blindness even with careful and meticulous conservative treatment of the individual episodes of uveitis. In about one-third of horses, both eyes are affected, often necessitating euthanasia. A link between ERU and leptospiral infection has been suspected for nearly 80 years. Vitreous lavage (vitrectomy) can preserve vision in affected eyes. After surgery, no further episodes of uveitis occur in up to more than 95% of operated eyes. With routine performance of vitrectomies, numerous vitreous samples could be used for further investigations. Intraocular anti-Leptospira antibody production was proven, leptospires could be cultured from the vitreous samples, and the LipL32 gene could be detected in the vitreous samples by PCR. Thus, there was convincing evidence of a chronic intraocular leptospiral infection, which can be eliminated most reliably by vitrectomy. Recently, it has been shown that the intraocular leptospires produce biofilm in the equine vitreous. Biofilm formation explains not only the success of vitrectomy, but also the survival of leptospires in the vitreous cavity for many years despite the presence of high intraocular antibody titers and immunocompetent cells, as well as the high tolerance to antibiotics.

1. Introduction

Equine recurrent uveitis (ERU) occurs in mules and horses and is a disease that has been known for a long time. From about the beginning of time-counting, ancient writings have described symptoms that are consistent with today’s definition of ERU. Since the end of the nineteenth century and the beginning of the twentieth century, more and more detailed descriptions of this disease have been published [1]. In earlier times, the working power of the horse was quite crucial for the survival of men [2]. Not only during war, but also for the cultivation of the fields, for the transport of people and freight as well as for serving as living motors in preindustrial times, people were dependent on horses and mules.

Without horses, the development of mankind would not have been possible to the extent that has been achieved in the past centuries. All the more the health maintenance of the horses was of paramount importance [3]. The recurrent and painful episodes of uveitis led to reduced performance and not infrequently to blindness and thus often to unserviceability of the affected horses. For this reason, equine recurrent uveitis has preoccupied many generations of owners and veterinarians [3, 4]. There are the most diverse historical treatment approaches and theories about the causes of this disease [3, 5].

Among many causes that had not been confirmed, wet pastures and flooding as well as heritability were discussed [5, 6, 7]. An infectious etiology has been suspected for over 100 years, although Leptospira spp. were not known at that time [3]. Since the first description of Weil’s disease in humans, “eye complications” were known to be associated with this disease [8]. A first description of leptospires was given in 1915 [9]. At that time, the identification of Leptospira spp. was made in Japan and Germany at approximately the same time [10].

After a link between leptospiral infections and uveitis had been established in human medicine, the Swiss ophthalmologist Gsell and coworkers studied aqueous humor from equine ERU eyes and described for the first time a link between ERU (then called “moon blindness” or “periodic ophthalmia”) and leptospiral infection [11]. Since then, there have been numerous investigations addressing the leptospiral etiology of ERU.

Because antibody detection in intraocular fluids was relatively common [11, 12, 13, 14, 15, 16, 17], but uveitis bouts typically do not become apparent until months or even years after the acute systemic infection [18, 19, 20, 21, 22], it was assumed that the infection was a trigger of ERU, but the bacteria were no longer present when the uveitis attacks started [18, 23]. In addition, a culture of Leptospira spp. from equine intraocular samples failed many times [12, 16, 22, 24, 25, 26, 27, 28]. For this reason, ERU has also been considered by some authors to be an “autoimmune” disease [29, 30, 31].

Different causes of uveitis can occur in horses just like in other species [30, 32, 33]. However, in equine uveitis associated with painful recurrent episodes causing the typical ocular changes, chronic intraocular leptospiral infection has been found to be the cause [34, 35]. Therefore, the term “ERU” will be used hereafter to refer to leptospiral-induced recurrent uveitis.

It was not until the routine use of vitrectomy (irrigation of the vitreous chamber) in horses [36, 37, 38] and the resulting ability to obtain intraocular specimens from eyes affected with ERU [39], that the importance of leptospiral etiology in ERU was confirmed [34, 35, 40, 41, 42, 43, 44].

Only recently it was recognized that recurrent uveitis in horses is a biofilm-mediated disease [45]. The ERU has many aspects that had raised questions and been incomprehensible before the discovery of biofilm formation of pathogenic Leptospira spp. in the vitreous chamber. However, knowing the characteristics of chronic and biofilm-associated infections, the pathogenesis of ERU can now be better understood [33].

2. Incidence and clinical course of ERU

Leptospiral-induced uveitis is not only in horses a late consequence of systemic infection [18, 34, 46], but also human leptospiral uveitis often occurs a long time after the acute infection [11, 46, 47, 48, 49, 50, 51]. A causal relationship between uveitis and a previous leptospiral infection is often difficult to recognize when uveitis occurs, because systemic leptospirosis is predominantly inapparent in horses [19, 52] and can also be inapparent in humans [53].

ERU affects quite a lot of horses. In the United States, where there are many leopard coat pattern horses (Appaloosas), it has been reported that up to 25% of horses are affected and lose vision in one or both eyes during the course of the disease [30]. However, in that study, leopard coat pattern uveitis (Section 4.), which accounts for a large proportion of affected horses in the United States, was also classified as ERU. In other studies, the percentage of horses affected with ERU ranges from 7 to 10% [54, 55, 56], with up to one-third of the horses suffering from the disease on both sides [3, 34, 57]. The attacks of uveitis in both eyes often do not start at the same time, but with a time delay of several months up to about 2 years [34].

The first episodes of uveitis are usually noticed in younger horses between 4 and 6 years of age [34]. More rarely, however, horses can still develop ERU up to over 20 years of age. Foals up to 6 months of age typically do not develop ERU. When uveitis occurs in foals younger than 6 months, it is typically septicemia-associated and bilateral, e.g., in the course of rhodococcosis [58, 59, 60, 61].

In ERU, recurrent episodes of uveitis occur in unpredictable intervals and oftentimes not, as the former term “periodic ophthalmia” suggests, periodically. The interval between episodes of uveitis can be less than 2 weeks and up to more than a year. In most cases, ERU episodes are associated with blepharospasm, epiphora, and photophobia, so the owner notices the eye disease and seeks veterinary advice. The severity of uveitis also varies greatly from horse to horse. Sometimes very mild episodes occur, which subside after 1–2 days. Other ERU attacks are so severe that after one or two attacks, the eye may already show significant and irreversible changes and in the worst case may even lose vision. In most horses, the clinically quiescent intervals between episodes of uveitis become shorter over time, and at the same time the uveitis bouts become more severe.

3. Clinical signs of ERU

Descriptions of the ophthalmologic findings in ERU have been given repeatedly and in broad agreement [3, 30, 32, 34, 62, 63, 64]. Acute attacks are usually painful or even very painful. Affected horses are depressed, show decreased appetite, can have a moderate rise in body temperature, severe blepharospasm, serous and later sero-mucous lacrimation, and more or less swollen eyelids. These symptoms, although typical, are not pathognomonic and can also occur with other eye lesions.

Ocular examination in horses is the easiest and most informative when a simple handheld (direct) ophthalmoscope with bright light source is used. The handheld ophthalmoscope can be used as a focal light source, magnifying glass, and slit lamp, and is most crucial for examining the posterior segment of the eye (posterior lens surface, vitreous cavity, and fundus). Since serum in horses is yellowish in color, aqueous humor and vitreous humor in acute uveitis (“leakage”) are also jaundiced. Ophthalmic examination typically reveals the following findings during an acute ERU episode:

• Reddening of the conjunctiva

• Low-grade diffuse corneal haziness

• Incipient circular vascularization of the cornea

• Jaundiced aqueous humor with positive Tyndall effect, usually also fibrin in the anterior chamber of the eye

• Miosis and only delayed and often incompletely medically achievable mydriasis

• Diffuse haziness of the vitreous humor

• Ocular hypotension (intraocular pressure often <10 mmHg)

In the inflammation-free interval, after mild ERU episodes and meticulous conservative treatment, sometimes no definite changes can be detected in early stages of the disease. However, when multiple ERU attacks have occurred, pathologic changes become increasingly apparent that are also evident during the clinically quiescent phase of the disease:

• Gradually increasing atrophy of the globe (if necessary, the inner anterior-posterior diameter can be measured by ultrasound; the difference is definite as from ≥2 mm side-to-side difference)

• Delayed pupillary response to light, drug-induced mydriasis also only achievable with delay

• In mydriasis, otherwise hidden posterior synechiae or iris residuals may be detected on the anterior lens capsule

• Diffuse vitreous opacification may still be recognizable in the inflammation-free interval (in some cases only evident by comparison with the other eye, and if the fundus on the diseased side is less clear compared with the other side)

• Vesicular cataract, typically in the periphery of the posterior lens capsule

• Dense vitreous deposits, initially visible only in mydriasis and typically located high dorsally close to the ciliary body; in the course of the disease, these deposits become more pronounced and can eventually also be seen in the center of the vitreous cavity, many times in combination with a murky yellowish discoloration of the liquefied vitreous

• In more advanced stages of the disease, moderate to severe bulbar atrophy or even phthisis, cataract, lens luxations, and retinal detachment may occur

In 3% of ERU cases, the inflammation occurs primarily in the posterior segment of the eye [65]. Hardly any pain is evident in these horses, and this form of ERU is sometimes detected only as an incidental finding during routine examinations or purchase examinations of horses. Only rarely do very observant owners notice a change in the fundus reflex of the diseased eye and call a veterinarian. In most cases, however, iritis occurs in the course of the disease, which then leads to the typical and easily recognizable pain symptoms. Depending on the changes that have already occurred in the posterior segment of the eye, the prognosis for preservation of vision is often guarded at this point. Sometimes these horses are not presented to the veterinarian until “sudden” blindness due to cataract formation or retinal detachment has occurred.

4. Differential diagnosis

A significant and strinkly common type of uveitis not caused by leptospires occurs in leopard coat pattern horses [65, 66]. This type of uveitis is strikingly common in leopard coat pattern horses. In contrast to ERU, leopard coat pattern uveitis progresses insidiously and does not present as recurrent painful episodes of uveitis. In the literature, it is therefore often referred to as “insidious uveitis,” but not distinguished from ERU. Other forms of uveitis may be phacogenic, traumatic, tumor-associated, septicemia-associated, or triggered by other infectious causes such as parasites (Micronema (syn. Halicephalobus) deletrix or Sertaria spp.) or, e.g., staphylococci [33]. In addition, a chronic iritis similar to Fuchs’ heterochromia iritis in humans occurs in horses [33, 67]. In most cases, all these forms of uveitis can be relatively clearly differentiated from ERU based on the clinical picture and/or the course of the disease (Table 1) [33].

	History	Typical ophthalmologic findings	Therapy	Prognosis
ERU	most times recurring episodes of painful uveitis attacks	see Section 3.	acute uveitis: topical atropine and dexamethasone, systemic nonsteroidal anti-inflammatory drugs, see Section 6.; quiet interval: see Section 7. and 8.	unfavorable, if exclusively conservative therapy is given, usually increasing cloudiness of the transparent media; good for prevention of recurrences with vitrectomy; good for permanent preservation of vision if vitrectomy is performed before ERU has caused irreversible intraocular damage
Leopard coat pattern uveitis	no obvious painful uveitis attacks, “suddenly” noticed impaired vision or cloudiness of the eye	lens pathology (initially or very early in the course of the disease), cataract formation, lens (sub-) luxation, sometimes glaucoma, sometimes atrophy of the globe	so far no etiologic therapy possible, only symptomatic treatment (anti-inflammatory, cyclosporine devices, combatting an elevated intraocular pressure)	unfavorable, often enucleation is required; if both eyes are affected, euthanasia may be indicated
Phacogenic uveitis	often mild uveitis, rarely severe uveitis attacks, sometimes presentation only because of cataract formation	lesions of the lens capsule, sometimes lens fragments in the anterior chamber or in the vitreous chamber, mild or moderate amount of fibrin in the anterior chamber, posterior synechiae (sometimes sealing defects in the anterior lens capsule)	conservative therapy see section 6.; phacoemulsification might be considered in selected cases	depending on the degree of leakage of lens proteins and the course of the disease good to unfavorable; phacoemulsification: guarded (risk of retinal detachment)
Traumatic uveitis	sudden onset of eyelid swelling, no previous uveitis attacks observed; in some cases: observed head and/or globe trauma	sometimes accompanying lesions of the eyelids or the cornea, often sero-hemorrhagic uveitis, sometimes hyphema or even hemophthalmus	if no corneal defects: see section 6., in case of corneal defects: no corticosteroids until the cornea is fluorescein-negative; if the fibrin and especially blood are not decreasing within about 10 days: injection of fibrinolytics (e.g., urokinase) or mechanical removal of the inflammatory products and the blood coagulum (e.g. using a small vitrectomy-cutter)	good if there has been no damage to lens and retina and no severe damage to cornea and sclera
Chronic iritis, similar to Fuchs’ heterochromic iritis in humans	no painful attacks, most often presented because of corneal cloudiness, rarely because of pigment loss of the iris	often endothelial precipitates and circumscribed depigmentations in the iris, increasing corneal edema, usually slow progression of the disease over years, glaucoma might finally occur	no etiologic, but only symptomatic therapy possible	guarded concerning long-term vision, sometimes enucleation required
Uveitis caused by severe keratitis or corneal infections (kerato-uveitis)	often very painful, increasing corneal opacification, no satisfying response to treatment	deep corneal ulcer or outlined round dense opacities, sometimes annular opacities (ulcus serpens)	corneal debridement, in case of severe weakening of the cornea and imminent rupture: suturing of a conjunctival flap into the lesion; in case of corneal rupture: corneal transplant	depending on the stage of the disease, the extent of the infection and the pretreatment (topical or systemic nonsteroidal drugs and especially corticosteroids worsen the prognosis) good to unfavorable
Uveitis accompanying septicemia	uveitis, most often in foals <6 months, rarely adult horses	typical acute uveitis, comparable to an acute ERU-attack, but often both eyes involved, presence of a severe general infection (e.g. rhodococcosis)	treatment of acute uveitis (Section 6.) and treatment of the septicemia	good if septicemia can be treated successfully and if conservative treatment achieves mydriasis; usually no further uveitis attacks occur
Uveitis caused by intraocular parasitic infections	chronic severe (kerato-) uveitis, not responding to therapy	cloudy cornea, depending on the causative parasite parasites might be visible in the anterior chamber or seen in the ultrasound examination	treatment of acute uveitis (section 6.), but usually uveitis is not responding; if parasites are visible in the anterior chamber: surgical removal may be discussed; enucleation might be indicated	for preserving vision: unfavorable; for preserving the globe: guarded to unfavorable
Uveitis caused by intraocular tumors	presentation because of (mild) uveitis, impaired vision and / or corneal or intraocular opacities	most often iris-melanoma, causing keratitis and sometimes glaucoma, other tumors occur less frequently (e.g. medulloepithelioma or malignant lymphoma); uveitis can be caused mechanically and immunologically	symptomatic treatment as long as possible, if the intraocular tumor causes chronic pain or if the tumor is going to spread enucleation is required	unfavorable for preserving vision and for preserving the globe
Endophthalmitis	disturbed general condition, fever, severe eyelid swelling	phlegmon of the eyelids, purulent epiphora, circular deep corneal vascularization, severe yellow-green corneal opacity	in case of very early stages: lavage of the vitreous cavity; otherwise enucleation	for vision: unfavorable; for preserving the globe: guarded to unfavorable

Table 1.

Different types of uveitis in horses, symptoms, therapy, and prognosis.

Sometimes recurrent keratitis is misinterpreted as ERU, as some types of keratitis can also cause painful with miosis and responds to the same conservative therapy as ERU. However, in keratitis cases, medical dilation of the miotic pupil results usually more rapidly and completely than in ERU. In recurrent keratitis, however, the changes that almost always are evident in ERU after several episodes of uveitis, even in the inflammation-free interval, are absent.

If an ocular disease is clinically not clearly assignable to an etiology (e.g., “recurrent keratitis” or “uveitis of unknown cause”), it is possible to take aqueous humor during the inflammation-free interval [33]. In horses, approximately 1 ml of aqueous humor can be safely collected and then used for laboratory tests [33, 68, 69, 70]. To investigate for the presence of ERU, testing for both anti-Leptospira antibodies and by PCR for, e.g., LipL32 is advisable [35, 71, 72, 73]. For scientific questions, a leptospiral culture can additionally be performed [34, 35]. Depending on the laboratory findings, a decision can then be made on the further course of action. In case of positive leptospiral findings, vitrectomy is indicated. With negative leptospiral laboratory findings, a leptospiral infection of the vitreous cavity can be excluded with a high probability. These horses would not benefit from vitrectomy—except to remove vitreous opacities that impair vision. In this case, however, a preoperative aqueous humor examination would be superfluous—just as in the case of unequivocal findings in terms of ERU.

For the detection of intraocular anti-Leptospira antibodies, the microscopic agglutination test (MAT) is used in most cases. The MAT is highly sensitive and specific when examining aqueous humor or vitreous samples [34, 35, 40]. In addition, other antibody tests can be used, which are either commercially available or available as in-house ELISA tests. Specific anti-Leptospira immunoglobulin class A (IgA) antibodies are particularly reliable for detecting intraocular leptospiral infection [72, 74]. Another well-suited test is the SNAP Lepto, which detects anti-LipL32 antibodies and is neither immunoglobulin-specific nor serovar-specific. It can be used for samples from different species. With its easy handling and the result visible within 10 minutes, this is a very useful test with a sensitivity and specificity comparable to MAT for intraocular specimens [73, 75]. In contrast to MAT, which is too unspecific for serum testing, SNAP Lepto is well qualified as a screening method even when serum is tested [71].

Antibody detections are equally reliable in vitreous and aqueous humor samples [70, 76, 77]. Both PCR and leptospiral culture are somewhat more reliable when testing vitreous humor samples compared with testing aqueous humor samples [34, 35, 78]. However, the collection of a vitreous sample is disproportionately risky and should be rejected for a preoperative diagnosis, because the aqueous humor analysis is overall very informative [33]. In rare cases, e.g., no anti-Leptospira antibodies are detectable in the aqueous humor, but at the same time the PCR yields a positive result. In routine diagnostics, culture has been largely replaced by the much faster and less expensive PCR.

If time is not an issue, but economic reasons have to be taken into account, a reasonable approach for the examination of aqueous humor samples is to first perform an on-site rapid test for the detection of anti-Leptospira antibodies. If this test is negative, the MAT can be commissioned externally if necessary. If the MAT is also negative, further antibody tests (e.g., specific in-house ELISA tests) and a PCR can be performed. The more laboratory tests are performed, the fewer “false negatives” can be expected, but the higher the costs for laboratory diagnostics will be.

5. Interpretation of intraocular antibodies

In eyes with a history of recurrent inflammations, but without clear evidence of ERU, and thus without aqueous or vitreous humor opacities, protein levels are typically not elevated. If protein levels in intraocular fluids are not elevated, leakage from the blood can be excluded. In these cases, even very low MAT titers are indicative of intraocular antibody production. The authors consider a MAT result of 1:50 as sufficient indication for vitrectomy in these cases. In eyes with obvious aqueous humor and vitreous opacities, however, the diagnosis of ERU is usually unambiguous even without aqueous humor examination. In cases of doubt, the Goldmann-Witmer coefficient can be used to differentiate leakage from intraocular antibody production [79]. It is crucial that not only the intraocular and the serum titer are evaluated, as it often could be read lately [80, 81, 82, 83, 84, 85, 86], but that—as described by Goldmann and Witmer—a reference value is determined both in the aqueous humor and in the serum. Any other antibody titer (e.g., tetanus) can be used as a reference value, provided that antibodies are present in the serum. Alternatively, the total IgG content or, if necessary, even the total protein content can be used as a reference value [33, 34].

6. Therapy of acute uveitis

Acute ERU is treated in the same way as any other equine uveitis [30, 32, 57, 58]. First of all, it is important to achieve mydriasis to avoid posterior synechiae and resulting cataract formation. Atropine is the drug of choice for this purpose and can be used as of 1–2% eye drops or eye ointment. Since the ophthalmic ointment adheres slightly better and acts more protracted, ointment is preferable, if available.

Atropine should initially be given several times daily or even hourly until the pupil dilates. Thereafter, the intervals can be adjusted to the pupil width and often considerably prolonged. Systemic side effects associated with the topical use of 1–2% atropine in horses do not play a significant role in the authors’ experience and after having treated thousands of horses over a 30-year period. Colic, e.g., due to an impaction or a meteorism, can occur in any hospitalized horse, not just ophthalmic patients. By feeding mash and monitoring the fecal consistency, an impaction can be detected early and countermeasures (e.g., administration of laxatives) can be taken to avoid more serious colic.

Apart from mydriasis, anti-inflammatory treatment is important. Topical application of ophthalmic ointments containing dexamethasone is particularly effective, provided the corneal epithelium is intact. If corneal defects are present, topical corticosteroids must not be given.

In addition, the administration of a nonsteroidal anti-inflammatory drug (NSAID) orally is indicated. Only in exceptional situations and in case of very significant diffuse vitreous opacification, systemic administration of prednisolone (1 mg/kg per os) for several days may be considered additionally. In these particularly severe cases with significant diffuse vitreous opacification, adjunctive therapy with a systemically given antibiotic, e.g., enrofloxacin [87], can also be performed, to eliminate at least part of the intraocular bacteria—even if this does not completely eliminate the infection [88].

Other measures accompanying the therapy are keeping the horse in a dark place and resting in the stall or just light exercise until the acute inflammation has subsided. If it is not possible to keep the horse in the dark, wearing a light absorbing mask can be considered.

6.1 Brief historical overview of the development of conservative treatment of uveitis in horses valid today (without treatment proposals that did not prove successful or were even questionable from an animal welfare point of view)

• Topical atropine has been recognized as an essential therapeutic mydriatic for equine uveitis as early as 1821 and has been considered a standard treatment for ERU in textbooks since 1842 [89]

• Topical cocaine has been recommended for the control of pain since the beginning of the last century [90]

• Salicylic acid preparations have been included among the treatment options for uveitis since 1922 [91]

• Corticosteroids have been used both parenterally [92] and topically [93] to treat uveitis since the middle of the last century

• In addition to eye drops and ointments, subconjunctival injections with corticosteroids [94, 95], later also with cocaine and atropine, were suggested to intensify the local effect

• Systemic administration of nonsteroidal anti-inflammatory drugs (NASIDs) has also been part of the standard treatment of acute uveitis in horses since their approval for veterinary use in the late 1970s (flunixin meglumine and phenylbutazone) [96, 97].

7. Vitrectomy during the quiet intervals

The most effective treatment for ERU is vitrectomy (removal of diseased vitreous and irrigation of the vitreous cavity). This surgery is performed exclusively in intervals without acute inflammation. Mechanical removal of the vitreous opacities caused by inflammation and accessible vitreous parts very reliably and permanently eliminates the leptospires in the biofilm. Postoperatively, up to 98% of eyes remain free of recurrences when surgery is performed properly [98]. If, exceptionally, further episodes of inflammation occur after surgery, a second vitrectomy can, if necessary, permanently eliminate the infection and prevent further episodes.

Vitrectomy as a vision-preserving procedure is a demanding surgery, having a relatively long learning curve. Prerequisites for successful performing vitrectomies are solid training, availability of for equine ophthalmo-surgery optimized, custom-made instrumentation and equipment as well as careful and intensive perioperative examination and conservative treatment. Any complication may have devastating consequences and can lead to blindness or even enucleation. Only rarely eyes that are already blind undergo surgery in order to prevent both future painful uveitis attacks and removal of the globe, which is cosmetically unsightly.

In order to perform vitrectomy with minimal complications, an experienced team (surgeon, sterile and nonsterile assistant, skilled anesthesiologist) is required, as well as expensive equipment and instruments specially adapted to the dimensions of the horse’s eye. For this reason, only a few specialized equine clinics perform vitrectomies to date. In clinics in which vitrectomy is performed as a routine procedure, it is a quick (total anesthesia time is about 40 minutes, the surgical instrument is in the eye <10 minutes) and relatively safe procedure with a very good prognosis [38, 39, 89, 99].

8. Other treatment options for ERU

Apart from vitrectomy, other treatment options have been described, of which two in particular are favored in recent publications. One consists of an intravitreal gentamicin injection. However, the recommended dosage for this purpose (4–6 mg) [80, 100, 101] is 3–4 times the drug concentration that was found to be “safe” with regard to retinal toxicity in experimental studies [102]. So far, there are no long-term results after these injections and the number of horses treated in this way is still limited. Surprisingly, gentamicin injection is not recommended exclusively for equine eyes with intraocular leptospiral infection; other forms of uveitis are also treated with this injection. Improvement after the intravitreal injection is also thought to result from the antibiotic gentamicin having immunomodulatory effects [103].

The second therapeutic option described since the turn of the millennium is the deep intra- or subscleral implantation of a cyclosporine device [104, 105, 106, 107]. These implants lead to less frequent and milder episodes of uveitis over a period of up to about 2 years. However, the uveitis does not stop completely, and if the effect wears off, a new implant may have to be inserted. Like gentamicin injection, implantation of cyclosporine devices is performed independently of leptospiral infection in the vitreous cavity. Only individual authors differentiate and use the implants exclusively when no leptospiral infection is detectable [86]. Attention should also be paid to the drug law in its current version, which currently prohibits the import of cyclosporine devices, at least in the EU [108].

However, neither gentamicin injection nor implantation of cyclosporine devices can remove the dense vitreous floaters that often lead to impaired vision. Over time, these deposits also often adhere to the posterior capsule of the lens and, just like extensive posterior synechiae, can lead to a cataract formation.

9. Results of the examination of intraocular specimens

In the literature, before the introduction of vitrectomy to the therapeutic measures against ERU, there were only very sporadic reports of cultural detection of leptospires in intraocular specimens from eyes affected with ERU [109, 110]. Numerous investigators failed to obtain cultural evidence of leptospires, casting doubt on chronic intraocular leptospiral infection. It was rather assumed that although leptospires somehow trigger ERU, the inflammations are not subsequently maintained by the presence of the pathogen [22, 24, 63, 68, 111, 112].

Vitrectomy was initially performed to remove vitreous opacities. The aim was to improve vision in the eyes affected by ERU [36, 37]. However, it soon became apparent that vitrectomy was surprisingly effective in preventing further episodes of uveitis. Therefore, more and more horses were sent to the clinic for vitrectomy.

It was only with the routine performance of vitrectomy that it had become possible to examine numerous vitreous samples from horses suffering from ERU. The peculiarity was that the samples were predominantly from eyes still able to see at an early stage of the disease. By collecting the first milliliters aspirated from the vitreous cavity before opening the intraocular infusion line, it was possible to use undiluted vitreous material for investigations. The results of these examinations, in turn, provided insights into which ocular findings were associated with leptospiral infection and which were not. It was also shown that the prognosis in terms of postoperative absence of recurrences was best when eyes with an intraocular leptospiral infection were treated by vitrectomy [98]. In this way, on the other hand, the indication for vitrectomy was optimized.

With careful assessment of the indication for vitrectomy and examination of undiluted vitreous specimens, MAT titers of 1:100 or higher were detected in 382 of 426 vitreous samples (90%) examined [34, 35]. In some MAT-negative specimens, specific anti-Leptospira antibodies (especially immunoglobulin class A) could be detected by an in-house ELISA [74]. Leptospires were culturally detected in 189 of the undiluted vitreous samples from 358 eyes (53%) affected with ERU [34, 35]. The positive cultures had been obtained only after optimization of the sampling technique and immediate sterile inoculation into a transport medium for mailing to a laboratory. The sensitivity of PCR is in between culture and antibody detection. In 70–77% of vitreous samples from eyes affected by ERU, the PCR result was positive [35, 73, 75, 113].

In Germany and neighboring countries, infections with leptospires of the serogroup Grippotyphosa are dominating, accounting for about 80% of intraocular infections in horses suffering from ERU. Infections with leptospires of the Australis serogroup account for about 13–14% of intraocular infections. Less frequently, leptospires of the serogroups Pomona, Sejroe, and Javanica were also detected in the vitreous samples from ERU eyes [34, 35, 114].

Vitreous samples obtained during vitrectomies from eyes affected by ERU were also used for histological and ultrastructural studies. It has been shown that the leptospires in the vitreous of eyes affected with ERU are surrounded by a homogeneous layer, which is lacking the leptospires from culture [115]. This homogeneous layer surrounding the leptospires could be extracellular matrix. In another study, in addition to phagocytosed leptospires, dense roundish structures were detected in vitreous material from eyes affected with ERU [116]. Some of these roundish structures had been phagocytosed, but others of these structures were so large that phagocytosis was impossible. These dense round structures could represent mature leptospiral biofilm constructs.

In 1971, Williams reported on immunologically mediated tissue damage in cases of equine uveitis [22]. However, autoimmune reactions that can be detected at the same time as the leptospiral infection [117, 118, 119, 120, 121] must be autoimmune phenomena accompanying the infection, since they cease as soon as the infection has been eliminated [33, 35]. Thus, there is no evidence of autoimmune disease following ERU.

10. Pathogenic Leptospira spp. and biofilm

Since many chronic infections are associated with biofilm formation, it has long been suspected that leptospires also form biofilm in vivo. In in vitro studies, biofilm formation of pathogenic Leptospira spp. was observed [122], and a detailed description of the three-dimensional structure of these biofilms was given [123]. The main focus with regard to in vivo biofilm formation was on small rodents, which are considered the main vectors of pathogenic leptospires and are chronic shedders. Following experimental infections, evidence of biofilm formation in the proximal renal tubules had been observed [124, 125]. Recently, there was also a description of in vivo biofilm formation in naturally infected rats [126]. At about the same time, biofilm formation of leptospires in vitreous samples from eyes affected with ERU could be demonstrated by immunohistochemistry (IHC) [45].

11. Characteristics of biofilm infections in ERU

Recurrent episodes of uveitis and the concomitant intraocular persistence of leptospiral infection over a long period of time meet the criteria of a biofilm infection [127, 128] very well [129].

The infection primarily affects the vitreous cavity. Possibly following the vitreous clearance, leptospires (more rarely) can also enter the anterior chamber of the eye and be detected there [33, 34, 35, 130, 131]. However, the infection obviously remains limited to the eye, there is no evidence of further spreading. As with other local biofilm infections, IgA antibodies are of particular importance in diagnostics [72, 132, 133, 134, 135].

One criterion of biofilm infections is the difficult cultural detection of the causative pathogen and ERU meets this criterion. Despite urgent suspicion of leptospiral infection in ERU (high intraocular antibody titers, intraocular antibody production), however, cultural detection of leptospires is demanding and often failed [24, 25, 26, 136].

In the vitreous of horses suffering from ERU, there are not only high antibody titers, but also immunocompetent cells (besides lymphocytes, especially plasma cells, macrophages, and granulocytes) [116, 137, 138]. The epithelium of the ciliary body shows many plasma cells in eyes affected by ERU [139]. In the area of the ciliary body and the iris root, even lymph follicles develop during the course of ERU, which contain B lymphocytes in the center [30, 140, 141]. Nevertheless, the immune system fails to eliminate the infection from the large vitreous chamber of the horse.

Leptospira spp. could be visualized in vitreous samples from eyes affected with ERU by immunohistochemistry (IHC). The infecting bacteria are bound to each other, and extracellular matrix could be demonstrated around and in between the bacteria [33, 45, 115, 142]. Leptospira spp. could be demonstrated in planktonic forms as well as in smaller and larger cell aggregates and in larger biofilm structures [33, 45].

Leptospires localized in the vitreous chamber show high tolerance to antibiotics. The first cultures were performed with samples from the entire lavage fluid collected during vitrectomy [41, 42, 44]. In the lavage fluid, the vitreous material was diluted about 10-fold and the lavage fluid contained 0.08 mg gentamicin/ml. This concentration had been shown to be 100 times the minimum inhibitory concentration (MIC) for WHO strains of pathogenic leptospires in vitro [143]. Cultures with these vitreous samples were less frequently positive than in later studies performed with undiluted vitreous samples [34, 35, 88], but nevertheless several culture sets eventually became positive after further inoculations and thus dilution of the antibiotic concentration [41, 42, 44].

Similar results were found in a study in which horses had been treated preoperatively intravenously with enrofloxacin. In the undiluted vitreous samples obtained at vitrectomy, the enrofloxacin content was above the MIC. Compared with the control group, in which more than 50% of the cultures were positive for pathogenic Leptospira spp., only 30% of the cultures in the group treated with enrofloxacin were positive. Thus, although the probability of a positive culture had been reduced with antibiotic treatment, reliable elimination of the infection was not achieved.

12. Discussion

ERU with persistent intraocular leptospiral infection over a long period of time meets all criteria of an infection associated with biofilm formation. The most likely route by which leptospires enter the vitreous cavity during acute systemic infection is by the fenestrated capillaries in the Pars plicata region of the ciliary body [30, 33]. In the healthy vitreous with its collagen fibers and viscosity, there are ideal conditions for the formation of leptospiral biofilm (Figure 1) [129].

Figure 1.

Schematic illustration of the discussed pathogenesis of equine recurrent uveitis (ERU) caused by a leptospiral biofilm infection in the vitreous chamber. Each uveitis bout leads to increasing damage to the intraocular structures. 1. Infection of horses with Leptospira spp. may occur on humid and muddy pastures or by drinking from standing waters. The bacteria can enter the blood stream via intact mucous membranes (e.g., oral cavity) or small skin lesions (e.g., on the legs). 2. Leptospira spp. most probably enter the vitreous chamber (VC) via the fenestrated capillaries of the pars plicata of the ciliary body (CB). 3. Leptospira spp. within the vitreous chamber attach to each other and to vitreous fibers, starting biofilm production. 4. Transmission electron microscopy using a vitreous sample from an ERU eye: Leptospira spp. are surrounded by extracellular matrix (reprint of [115] courtesy of Schluetersche specialized media GmbH, Hanover, Germany). 5. Most Leptospira spp. are protected within the biofilm, single planktonic bacteria are in the vitreous chamber. 6. Vitreous samples from ERU eyes, containing visible inflammatory products (“vitreous floaters”); the yellow color indicates increased permeability of the blood-ocular barrier. 7. Threshold exceeded, immune privilege of the eye temporarily suspended, clinically apparent uveitis bout (left: Epiphora and blepharospasm; right: Much fibrin in the anterior chamber).

The vitreous body is 98% water and contains a collagen fiber scaffold. It has been shown that plant fibers in rice fields are important sites for biofilms [144]. The vitreous fibers [138] might also serve as “surfaces” to which Leptospira may adhere and start biofilm production. Furthermore, viscous media promote biofilm production of Leptospira spp. [145], and healthy vitreous humor is such a viscous substance. With the collagen fiber scaffold and viscous consistency, the vitreous thus represents an ideal medium for biofilm formation of Leptospira spp. [33, 45, 129, 145].

Another factor to consider is that the vitreous cavity of the horse has a volume of approximately 28 ml, making it a large immunologic niche [34]. In addition, there is the immune privilege of the eye [146, 147], which effectively suppresses the immune defense. In this way, pathogenic Leptospira spp. can remain clinically unnoticed in the eye for a long time. The latency period can be many months or several years. It probably varies with individual factors of the host, the amount of Leptospira spp. in the vitreous, and possibly the leptospiral serovar involved.

Only after months or years, when a threshold is exceeded due to gradual multiplication of the leptospires and increase of immune reactions despite the ocular immune privilege, a uveitis attack with disturbance of the blood-aqueous barrier or blood-ocular barrier becomes apparent [33, 34]. The immune response that occurs in conjunction with the inflammation likely results in the elimination of some planktonic bacteria. Other bacteria in the biofilm outlast the inflammatory bout. After the inflammation subsides under antiphlogistic treatment and with the help of intraocular immunosuppressive mechanisms, a clinically apparently inflammation-free interval occurs, which, however, does not represent a totally quiescent phase immunologically [137].

There are reports, and some own experiences seem to support this, that episodes of uveitis can be triggered by exposure to stressful situations (e.g., competitions, long-distance transport, change of stables, general anesthesia and major surgery). It is conceivable that endogenous cortisol release in stressful situations further reduces the immune defense in the eye (in addition to the ocular immune privilege). This in turn might increase the number of planktonic Leptospira spp. in the vitreous cavity after a stress situation and lead to contact with the uvea—which then causes an exaggerated immune reaction resulting in a uveitis attack.

A gradual spread of biofilm structures in the vitreous cavity could explain that ERU episodes occur at shorter intervals and become more severe over time. In addition, there are immune reactions that fail to eliminate the leptospires but may result in damage to the ocular structures adjacent to the vitreous chamber. One example is neutrophil extracellular traps (NETs), which have been detected in vitreous samples from eyes affected by ERU [148]. These NETs are formed by granulocytes to remove pathogens too large for phagocytosis [149]. A disadvantage of the formation of NETs is that tissue-damaging substances are also secreted, which in turn promote an inflammatory reaction of the surrounding tissue [150, 151], which in ERU cases is the uvea.

The high MAT titers in eyes affected by ERU certainly also play a crucial role in the course of the disease, as they promote agglutination of planktonic leptospires. However, since complete elimination of the bacteria is usually not possible, this agglutination can also be the starting point for new biofilm formation. During agglutination, leptospiral aggregates are formed, extracellular matrix is produced after surface contact of bacteria with each other, and thus new biofilm structures can be built. In this way, the agglutinating antibodies could accelerate the biofilm formation of pathogenic Leptospira spp. [33].

High levels of serum amyloid A (SAA) [152] and the formation of AA amyloid [153, 154] were detected in intraocular samples from eyes affected with ERU. The formation of amyloid is a good explanation for the fact that the dense vitreous floaters in ERU fail to resolve, but instead increase as the disease progresses. Besides the collagen fibers of the vitreous scaffold, the NETs and the amyloid fibers provide additional fiber structures that could be used for biofilm formation. The formation of NETs and biofilm promote each other [155, 156]. Similar to what has been described for otitis media [157], these numerous fibers could be incorporated into the biofilm and help to reinforce the biofilm scaffold, so that therapeutically only mechanical removal is promising.

With knowledge of the successful cultivation of leptospires from vitreous specimens that contained an active level of gentamicin or enrofloxacin above the MIC, it is questionable whether intraocular gentamicin injections, which are performed therapeutically by some veterinarians, provide lasting success. Biofilms can increase tolerance to antibiotics up to 1000-fold compared with planktonic bacteria [158, 159]. The described improvement of eyes suffering from ERU after gentamicin injection could be due to the fact that planktonic bacteria are eliminated. However, it is questionable whether the bacteria in the biofilm can really be eliminated by the injection. It could also be that the structure and composition of the biofilm change accordingly, so that the bacteria survive protected in the biofilm and then lead to ERU relapses again after some time. With the therapeutically used cyclosporin-devices, spread of the leptospiral biofilm in the vitreous cavity could even be favored, since immune reactions of the host, including those directed against the bacterial pathogen, are suppressed.

In vivo biofilm formation has also been described for other spirochetes. In human medicine, for example, chronic Lyme disease with its various organ manifestations plays an important role [160, 161]. In patients with Lyme disease, in vivo biofilm formation was shown to be associated with the long-term persistence of Borrelia spp. [162], and biofilms were found to contain both Borrelia spp. and Chlamydiae [163]. For example, alginates have been found in biofilms of Borrelia [164]. Alginates induce a distinct immune response [165] and result in the biofilm being more pathogenic than the planktonic bacteria. For lymphocytoma [166] and Alzheimer’s disease [167, 168], there are detailed descriptions of biofilm formation and indications for improved treatment options. In Alzheimer’s disease, Borrelia bacteria in planktonic form do not appear to cause noticeable harm. Here, too, it is the biofilms that create the pathology [167]. Biofilm formation and approaches to improve therapy have also been demonstrated following experimental Borrelia infections of mice as a model for Lyme disease [169].

The composition of leptospiral biofilms in the vitreous cavity in ERU is still largely unknown. Neither alginates nor curli fibers (bacterial amyloid) could be detected in the in vitro Leptospira biofilms [123]. The in vitro biofilms of Leptospira spp. consisted predominantly of extracellular DNA. However, the composition of in vivo biofilms of leptospires could be quite different [170]. It is possible that further analysis of the leptospiral biofilms in the vitreous cavity of horses suffering from ERU may provide further information on how to disperse these biofilms in a manner that is as tissue (retina, lens capsule) compatible as possible. This could provide new insights for the treatment of other biofilm-associated infections that are also relevant to human medicine.

13. Conclusions

ERU is a spontaneously occurring intraocular leptospiral biofilm infection. For centuries, only symptomatic conservative treatment was possible, which has become increasingly effective with the availability of modern anti-inflammatory drugs. However, even the most potent anti-inflammatory treatment could not prevent recurrences of uveitis, which led to gradual damage and even destruction of the affected globe. It was not until the introduction of vitrectomy in equine ophthalmology that causative therapy had become possible. Samples containing leptospiral biofilm can easily be collected in the course of therapeutic vitrectomy. Not only can these samples be used for laboratory diagnostics regarding intraocular leptospiral infection, but further studies can be performed on the composition of the biofilm. There could be significant differences between the composition of the biofilm formed in vitro and that formed in vivo, as host tissues (here: vitreous material and collagen fibrils) and interactions with the host immune system (e.g., agglutinating antibodies, macrophages, granulocytes, NETs, fibrin, and amyloid) influence the composition of the biofilm. ERU provides possibilities for investigation of an in vivo biofilm infection without the need for animal experiments and, thus, could serve as a naturally occurring entity for further research.