Background
“Corneal infiltrates” and “corneal ulcers” are one of the most common emergency issues, both in Acute and Emergency Eye Division (A&E) and the corneal specialists’ office [
1]. Although it is of the highest importance to properly assess the origin and implement the right treatment, direct diagnosis is far from straightforward. Dalmon et al. proved that even experienced corneal specialists can properly identify the etiology of corneal ulcers in 66% of cases based solely on anterior segment slit-lamp photographs [
2]. Similar, or even higher, uncertainty about the etiology arises when the special type of corneal infiltrate termed “ring infiltrate” is present.
Corneal ring infiltrate (CRI), by definition, is a ring-shaped stromal infiltrate of 360 degrees, circumferential to the limbus, typically leaving a clear zone from it. Ring infiltrates may originate from both infectious and sterile processes [
3]. The existing body of evidence associates CRI mainly with Acanthamoeba keratitis and fungal keratitis in their late stages, but the list of potential origins substantially exceeds the aforementioned causes [
4].
Despite the well-established definition and morphology, several questions about appropriate differential diagnosis and management of CRI arise. If corneal scraping is the gold standard ancillary test for each CRI differentiation? Which laboratory tests should be performed? Can anterior segment optical coherence tomography (OCT), in vivo confocal microscopy (IVCM), general physical examination, and blood tests also be helpful in diagnosis? Do these infiltrates appear when patients are symptomatic from a systemic process or are they the first symptoms of the disease? How does the prevalence of underlying disease change the perspective of ancillary testing performed when CRI is present. Does the depth of the ring suggest etiology and affect final visual outcome?
This review aims to provide a comprehensive analysis of potential origins of corneal ring infiltrate, characteristic of their morphological features, ancillary testing useful in differential diagnosis, the importance of a medical history, and consideration of general diseases and medications. Moreover, we attempt to differentiate between corneal ring infiltrate and other similar conditions such as peripheral ulcerative keratitis.
Nomenclature
The nomenclature of “corneal ring infiltrates” is ambiguous. In the literature several descriptions might be found: “sterile corneal ring infiltrate”, “infectious ring infiltrate”, “non-infectious ring infiltrate”, “sterile ring”, “immunological ring infiltrate”, “Wessely ring”, “ring abscess”. To clarify “infectious ring infiltrate” refers to a ring of microbial origin, where bacterial/viruses particles are found in the scraped material of the ring. It also correlates with clinical symptoms of infection (hyperemia, pus, anterior chamber activation, decreased vision). “Ring abscess” depicts an infectious ring, particularly suppurative ones. “Immunologic ring” is equivalent to “sterile” and “non-infectious” ring. “Wessely ring” is a type of sterile infiltrate based on type 3 hypersensitivity response marked by antibodies [
5].
Of note, peripheral ulcerative keratitis (PUK) is often linked to ring infiltrates with similar pathophysiology of immunological complexes activating complement, but collagenase (metalloproteinases) activation leading to corneal stroma thinning is required in this condition [
6].
By definition, the ring affects 360 degrees of the peripheral cornea leaving a clear margin from the limbus. In general, the ring appears more often in the peripheral cornea, where the number of antigen-presenting cells surpasses the central cornea [
7]. However, at the disease onset single peripheral infiltrates might be present, then coalescing in a ring-like one [
8]. Although corneal ring infiltrates originate from corneal stroma activation, the endothelium might also be affected. Endothelial rings occur more often in fungal than bacterial keratitis [
9]. Concomitant deprived epithelium also enforces infectious origin.
Pathophysiology
Cornea is a transparent structure, deprived from blood and lymphatic vessels. The ring forms typically close to the limbus, close to the conjunctival vessels, where the ratio of antigen-presenting cells (APC) exceeds the density of APC in the corneal center similarly to C1 concentration [
7]. Bacterial toxins damage host tissues by activating alternate complement pathway via properdin and C3 stimulation. Chemotactic agents attract polymorphonuclear leukocytes into the cornea [
10]. No immunoglobulins (of any class) are found in histopathological samples of infectious rings [
11]. Infectious rings form within 24–72 h on average. Immunological rings origin from two ways of complement activation: a) antibodies reacting with corneal particles from damaged tissue- type 3 hypersensitivity (Wessely ring) or b) antibody-independent complement activation. Of note, infectious inflammation might also lead to purely immunological corneal ring infiltrate in late mechanism of hypersensitivity (via release of microbial endotoxins triggering properdin and finally complement activation on antibody-independent way) [
12]. Furthermore, exposure of compromised corneal epithelium to saprophytic flora (antigen) might contribute to sterile ring infiltrate formation in contact-lens overwear, corneal foreign body, corneal burns and intoxication, recurrent corneal erosion syndrome, topical anesthetic abuse, post-surface refractive surgery, and in corneal cross-linking.
Etiology
Tables
1 and
2 depict the potential origins of infectious and non-infectious ring infiltrates. It is important to distinguish the origin from infectious and non-infectious etiologies. Additionally, we need to reiterate that microbial causes may also trigger an immunological ring. Still, the mainstay of primary differentiation is based on clinical signs and symptoms. Patients presenting with decreased vision, pain, pus in conjunctival sac, anterior chamber reaction, epithelial defect, corneal melting and less common- necrosis raise a high suspicion of infectious ulcer [
4,
13]. They occur 24–48 h after inoculation and are visible clinically after 2–3 days [
14]. Untreated infectious rings lead to centrifugal expansion deteriorating corneal stability.
Pseudomonas aeruginosa has been proven to impair collagen structure leading to descemetocele or melt instantly (within 7 days) [
15]. Sterile rings typically manifest with eye irritation, mild photophobia and decreased vision when corneal structure is damaged (thinning and finally scaring process present) [
16]. Immunologic rings slowly expand centripetally and eventually fade. Both etiologies may leave significant corneal thinning and neovascularization. Wessely’s rings form within 10–14 days from the trigger, but may appear within 1–5 days with previous exposure to the antigen [
17].
Acanthamoeba and
Microsporidium are the exceptions where immunological ring forms as a result of inappropriate treatment, with > 16 days in Acanthamoeba and even 2 months in Microsporidial keratitis [
18,
19].
Table 1
Infectious etiology of corneal ring infiltrate (CRI)
Bacteria- Gram-positive | Staphylococcus Streptococcus Bacillus cereus Listeria monocytogenes Nocardia | - 4% prevalence - early presence (24-48 h from inoculation) - associated hypopyon, hyperemia, epithelial defect, decreased vision) - mainly Gram-negative and mixed infections |
Bacteria- Gram-negative | Pseudomonas Moraxella Serratia Neisseria Klebsiella Escherichia Proteus Mycobacterium Capnocytophaga Microsporidium |
Fungi | Aspergillus fumigatus Aspergillus flavus Aspergillus niger Fusarium Acremonium | - 1–25% prevalence - yellow or creamy-white - coexistence of patchy, stromal infiltrate - long-lasting (presents till day 7, persists to at least 1 month) |
Viruses | Herpes simplex Varicella zoster | |
Acanthamoeba | Various Acanthamoeba subspecies | - 30% of cases - presents late (on average 7–14 days from infections’ origin) - greyish - 9–11 times more often than in fungal origin |
others | Infectious crystalline keratopathy (Streptococcus mitus) | |
Table 2
Non-infectious etiology of corneal ring infiltrate
General diseases | Rheumatoid arthritis Cryoglobulinemia Multiple myeloma and other hipergammaglobulinemias Amyloidosis |
Ocular diseases | Posterior polymorphous dystrophy |
Medications- topical | NSAIDS, topical anaestetics’ abuse |
Medications- general | Perifosine |
Post-laser | Collagen cross-linking Post-refractive surgery (excimer photoablation) |
Foreign body | Corneal foreign body Corneal intacts Corneal burns Insects hair/sting |
Toxins | Coral toxins Insects’ toxins Endotoxins (secondary to endophthalmitis) |
Trauma | Contact-lens wear Recurrent corneal erosions |
Others | Behcet’s disease |
Epidemiology
There is no data on exact prevalence and incidence of the ring infiltrate. However, when the general prevalence of diseases associated with CRI (especially keratitis) is considered some statistics might be indirectly calculated. Mascarenhas reported that 30% of Acanthamoeba keratitis 5% of fungal and 4% bacterial infiltrates present with corneal ring infiltrate at some stage [
4]. Bharathi proved the presence of CRI in 1.4% of fungal infections [
20]. Several reports of non-infectious ring infiltrates in CL-wearers show 0.5–6% annual risk of symptomatic sterile infiltrate [
21,
22]. Post cross-linking CRI were diagnosed in 2–8% of cases [
23‐
25].
Considering that 5% of global population wears contact lenses, they prevail among sterile CRI origins [
26]. A study on the prevalence of keratitis showed 0.15% of population affected, with 44% of viral (0.065%), 46% of bacterial origin (0.068%), 10% of fungal (0.015%) in a study by Cao resulting bacterial as a second most common etiology of CRI [
27].
Wearing contact lenses increases the incidence of infectious keratitis, sterile corneal ulcer (including ring infiltrate, also called “CLACI”- contact lens associated corneal infiltrates), contact lens peripheral ulcer (CLPU) and CL-induced acute red eye (ang. contact lens acute red eye- CLARE) [
28‐
31]. CLPU and CLARE are mainly associated with extended CL wear. CLPU is manifested with peripheral epithelial full-thickness, regular is shape lesions with co-exisiting corneal stromal infiltrate [
28]. CLARE is defined by sudden onset of painful red eye, typically early in the morning, where slit-lamp findings comprise: conjunctival and limbal redness and corneal infiltrate [
32].
All abovementioned might be manifested with ring infiltrate. Differentiation relies on clinical criteria. Infectious infiltrates resemble previously described non-sterile infiltrates. Gram (-) bacteria are found more often in CL-related ulcers, with Pseudomonas aeruginosa being the most common origin, with poor final visual outcome [
31]. Of note, atypical cases of sterile, CL-linked, infiltrates are described [
3]. Tabatabaei reported bilateral deep infiltrate affecting corneal epithelium (epithelial defect and oedema), anterior and deep stroma (dense ring infiltrate) coexisting with pain, redness and discharge. However, difference in epidemiology was described, with higher total incidence in CL-wearers, particularly of Acanthamoeba and fungal origin compared to non-CL-wearers [
32]. Of note, factors increasing prevalence of infectious keratitis in CL-wearers are: extended wear, swimming in CL, handwashing, overnight use, low hygiene level, short term of applying CL (< 6 months), male gender, onset in winter, young age, poor hygiene, internet purchase and smoking [
31,
32].
Sterile infiltrates in contact lens wearers are noticed more often in hydrogel than silicone lenses. Whereas extended wear increases the risk with odds ratio of 4–5 compared to daily wear, numerous risk factors are coined for all types of lenses: omitted/infrequent disinfection, bacterial contamination of the storage case, initial period of adaptation to CL, prior inflammation related to lenses wearing, smoking, age (< 25 years ld and > 50 years old), limbal redness, corneal staining, high ametropia [
31]. Additionally, toxic corneal staining poses a threat of sterile infiltrate also in daily disposable wearers with continuity of risk correlated to area of epithelium defects.
Sterile ring infiltrates in CL wearers is typically located in the superficial part of anterior stroma and resolves within 1–3 weeks. In the majority of cases no final visual acuity deprivation is noted [
29].
In summary, the most common infectious origin of CRI is bacterial keratitis and non-infectious cause is contact lens related.
Conclusions
The corneal ring infiltrate is widely linked to Acanthamoeba infection. However, the literature shows a wide variety of potential causative agents, from bacteria, fungi, viruses through immunological processes (drugs toxicity, CL wear, general diseases) to foreign bodies and trauma. Unfortunately, no pathognomonic symptoms nor signs related to CRI origin have been found so far. The precise differentiation remains challenging even though new technologies are applied (As-OCT, IVCM, artificial intelligence). An existing large body of evidence suggest that corneal ring infiltrate should be scraped and treat empirically with anti-infectious agents until proven otherwise. Laboratory tests are ordered in scrapes negative patients with no co-existing risk factors for sterile infiltrates. Infectious rings, either confirmed in corneal sample result or suspected clinically and improving on proper medication, need to be treated as microbial keratitis. Sterile infiltrates fade within time in general with no treatment, but topical steroids facilitate resolution of Wessely’s ring. Non-infectious etiology recurs more often, resulting in scars and melting.
Further research is of need to guide the clinicians with most probable etiology and proper management of CRI.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.