ASSESSMENT AND MANAGEMENT OF BACTERIAL KERATITIS IN CONTACT LENS WEARERS

Joseph P. Shovlin, OD, FAAO, Jared W. Schneider, BS, Helen J. Chandoha, OD

Course List | Take This Exam | CE Home | Optometry's Home

Introduction

Organisms Causing Bacterial Keratitis

Table 1. Regional variation in the organisms cultured from corneal scrapes.²

	Staphylococcus	Streptococcus	Pseudomonas	Other Gram Negatives	Fungi
New York	49%	9%	8%	22%	3%
Florida	16%	8%	19%	9%	35%

Table 2. Variation in organism cultured from corneal scrapes at Moorfield’s Eye Hospital from 1985 to 1995. ²

	Staphylococcus sp.	Streptococcus sp.	Pseudomonas	Other Gram Negatives	Acanthameoba	Fungi
1985	29%	22%	19%	25%	0%	13%
1995	34%	13%	21%	12%	13%	6%

Specific Risk Factors for Bacterial Keratitis

Table 3. Risk Factors for Bacterial Keratitis (adapted from: Matoba, AY: Infectious Keratitis; Focal Points-Clinical Modules for Ophthalmologists. September, 1992)

Exogenous	Contact lenses, contaminated cases and solutions, foreign body, chemical and thermal injury, previous ocular surgery, medicametosa, contaminated medications and make-up.
Adnexal Dysfunction	Trichiasis, abnormal lid anatomy and function, tear deficiencies, conjunctivitis, neuropathy of cranial nerve(s) III, V, and VII, blepharitis, canaliculitis.
Corneal Abnormalities	Hypesthesia, bullous keratopathy, erosive disorders, viral keratitis.
Systemic Disease	Diabetes mellitus, malnutrition or respirator dependence, collagen vascular disorders, substance abuse, exfoliative skin disease, immunocompromised patient, atopic dermatitis, vitamin A or B deficiency.
Immunosuppressive Therapy	Systemic corticosteroids, topical immunosuppressive agents, systemic chemotherapy for malignancy, or transplant.

Bacterial Keratitis Associated With Contact Lens Wear

Pathogenesis of Bacterial Keratitis

An intact corneal epithelium provides an effective barrier against most microorganisms; the exceptions are uncommon ocular pathogens.¹ N. gonorrhoeae, Corynebacterium diphtheriae, Listeria, and Hemophilius aeggyptius are unique in their potential ability to infect the cornea without disturbing the epthelium.¹

Most often microorganisms gain access and invade the cornea through epithelial defects. The site of entry can be minor, such as a mascara brush injury or a small abrasion associated with contact lenses wear. Once the microorganisms gain access to corneal structure, the primary determinant of the disease severity is the organism's virulence.

Virulence is best characterized by an organism’s ability to induce chemotaxis and release proteolytic enzymes. Pathogenicity is directly related to its ability to adhere to the edge of the epithelial defect and initiate infection. Bacteria such as S. aureus, S. pneumonia, and P. aeruginosa are responsible for a high number of infections in contact lens wear due to traits such as pili and glycocalyx that facilitate adherence and produce aggregates that resist phagocytosis.¹

Bacterial enzymes facilitate corneal tissue necrosis and aid in penetration deep into the stroma. Multiple toxins are released by Staphylococcus (alpha, beta, gamma, and delta) and Pseudomonas (A, B, and C) species in addition to proteases, coagulases, collagenases, and lipases.¹ These enzymes and toxins often remain after the death of the pathogen and can continue to cause tissue damage.¹

Exotoxins produced by both gram positive and gram negative bacteria and endotoxins contained in the cell wall of gram negative bacteria play an important role in the virulence of each species. S. aureus, a common pathogen implicated in bacterial keratitis, produces over two dozen biologically active substances and toxins with coagulase-positive strains being the most pathogenic.

Pseudomonas species produce exotoxin A which inhibits cellular protein synthesis and a corneal destroying proteoglyanase that breaks down ground substances around collagen. Endotoxins are released when an organism dies and can cause an inflammatory corneal ring. These corneal rings are thought to be antigen-antibody precipitates.¹

Host enzymes have a role in the destructive process of bacterial keratitis. Particularly, the pathogenesis of Pseudomonas corneal infections is mediated by host-derived enzymes. Chemotactic substances released during corneal invasion initiate polymorphonuclear (PMN) leukocyte migration from the limbal area. PMNs are able to destroy microbes through degranulation or respiratory burst. These defenses are important in microbial destruction but are also associated with destruction of the corneal collagen matrix and ground substances. The amount of visual loss produced by microbial keratitis is directly related to the extent of inflammatory cell infiltration, cell death, release of proteolytic enzymes, and damage to the endothelium.¹

Clinical Features of Bacterial Keratitis

The symptoms of infectious keratitis include severe pain, redness, decreased vision, and photophobia. Signs include significant lid edema and reactive ptosis, conjunctival and ciliary injection, discharge, papillary response, stromal infiltration, surrounding tissue edema, epithelial defect, anterior chamber reaction, cellular debris in the tear meniscus, and even hypopyon.

These signs and symptoms are influenced by prior corneal status and therapy, the virulence and pathogenicity of the organisms, the length of the infection, and the host immune response to the infection. However, there are no specific symptoms or definitive biomicroscopic features that distinguish the responsible organisms.

Epithelial defects with mucopurulent exudates adherent to the ulcer, focal stromal suppuration, diffuse stromal infiltrates, radiating folds in Descemet’s membrane and an anterior chamber reaction with or without a hypopyon are strongly suggestive of bacterial infection.¹

In patients with previously altered corneas, the signs and symptoms of bacterial keratitis are different. Corneas with prior inflammation, structural changes, and presence of an iritis make it difficult to determine if a bacterial infection is present. In patients such as these, infection should be suspected if there is increased pain, redness, enlargement of epithelial and stromal ulceration, or an increase in anterior segment inflammation.¹

Figure 1. A sterile mid-peripheral infiltrate in a disposable lens wearer.

Figure 2. A sterile ulceration from crack cocaine use resulting in a chemical burn and neurotrophic cornea.

Table 4. Distinctive Signs of Bacterial Keratitis.

	Typical	Atypical
Epithelium	Ulcerated	Intact
Stromal Inflammation	Suppurative	Non-suppurative
Site of Inflammation	Focal, diffuse	Multifocal

Differential Diagnosis of Ulcerative Lesions

Infectious Etiologies

• Bacterial: The most common of infectious etiologies
• Fungal: Should be considered after trauma involving organic or vegetative materials. Infiltrates are often present with feather borders and satellite lesions. Hypopyon formation is possible but not diagnostic.
• Acanthamoeba and other protozoan: Extremely painful with radial nerve infiltrates, sometimes producing pseudodendrites. Late stages of infections exhibit a ring stromal infiltrate. Patients often have a history of poor contact lens care, hot tub use, or swimming with contact lenses.
• Herpes simplex virus: Corneal epithelial dendrites or geographic ulcers are possible. Chronic herpes simplex keratitis can easily develop into bacterial superinfections.

Figure 3. A filamentous fungal infection with satellite lesions caused by Fusarium.

Figure 4. A protozoan infection caused by Microsporidia following a bee sting to the cornea.

Non-Infectious Etiologies

• Sterile ulcer: Associated with keratoconjunctivitis sicca, collagen-vascular diseases, vernal keratoconjunctivitis, neurotrophic keratopathy, crack cocaine use (Figure 2), and vitamin A deficiency.
• Staphylococcal hypersensitivity: Characterized by multiple peripheral corneal infiltrates with a clear space between the infiltrates and the limbus. Most often bilateral and associated with blepharitis.
• Sterile corneal infiltrates from an immune response to contact lenses and solutions.
• Contact lens peripheral corneal ulcers: Found in contact lens wearers, these peripheral inflammatory lesions resemble infectious keratitis. In addition to their classic circular appearance, sometimes these infiltrates will have a "tail" effect.

Sterile vs. Infectious Infiltrates

The differential diagnosis of infiltrative keratitis (Table 5) is extremely important in order to provide appropriate treatment and reduce the potential for corneal morbidity. Sterile corneal infiltrates are generally small (less than 1 to 2 mm) and are often located in the mid-peripheral and peripheral cornea. They represent aggregates of white blood cells, mostly lymphocytes and PMNs that have migrated from the vasculature near the corneal limbus into the precorneal tear film.³ Generally there is little to no epithelial defect or anterior chamber reaction.

The patient may present with some photophobia, but little to no pain. However, there may be a foreign body sensation. In contact lens wearers, these complications are most likely due to chronic hypoxia, preservative hypersensitivity, or perhaps colonization of the lens surface and antigen exposure to the cornea of low virulent gram negative bacteria.⁴

Corneal infiltrates/ulcers from bacterial infection typically show overlying epithelial defects most often larger than the area of initial infiltration. There is also underlying loss of Bowman’s layer and necrosis of the stroma.

These infiltrates are generally more centrally located and are greater than 1.5 to 2.0 mm in diameter. There can be a full-thickness stromal infiltration producing corneal edema and possible folds in Descemet’s membrane. The bulbar conjunctiva will be more diffusely injected than is expected in sterile reactions. Cells and flare are noted in the anterior chamber and a hypopyon can develop, but these manifestations are not diagnostic for bacterial infection.

The patient will present with significantly more pain, photophobia, sometimes decreased vision, and increased foreign body sensation. The eyelids will be chemotic, often with a reactive ptosis. Mucopurulent discharge and debris in the tear film are commonly associated with microbial keratitis secondary to bacterial infection.

Table 5. Differential Diagnosis of Sterile Infiltrates versus Infectious Corneal Ulcers.

Characteristics	Sterile	Infectious
Location	Generally Peripheral	Central
Size	Small (0.1 to 1.5 mm)	Large (over 2.0 mm)
Epithelial Defect	Usually little or no staining	Staining present
Conjunctival Injection	Localized	Diffuse
Anterior Chamber Reaction	Absent or mild reaction	Mild to severe
Discharge	Absent or serous	Mucopurulent
Pain	Absent or mild	Moderate to severe
Photophobia	Mild or absent	Present
Culture	Negative, unless contaminated	Positive, unless false negative

Figure 5: Aggregates of white blood cells in sterile infiltrates. (Courtesy of L. Catania.)

Figure 6: Ulcerative keratitis showing loss of tissue and infiltration. (Courtesy of L. Catania.)

Documenting and Grading Corneal Ulcers

The initial presentation of the ulcer should be well documented. Important ulcer characteristics that should be noted for comparison during future examinations include:

• Location, size, and shape of the epithelial and stromal defect,
• Suppuration and edema,
• Amount of anterior chamber reaction,
• Any scleral involvement.

A grading system (Table 6) is beneficial in assessing ulcer severity and can provide a guideline for therapy. An ulcer with focal superficial suppuration can be classified as mild; a large superficial inflammation that is limited to the anterior two-thirds of the cornea would be considered moderate; and an extensive area involving the posterior one-third of the stroma with scleral suppuration or threatened perforation is graded severe.¹

Table 6. Grading scheme for corneal ulcers.

	Mild	Moderate	Severe
Size of defect (mm)	Less than 2	2 to 5	Over 5
Depth of ulcer (%)	Less than 20	20 to 50	Over 50
Infiltrate	Superficial	Dense, mid-stromal	Dense, past mid-stroma
Sclera	Not involved	Not involved	May be involved

Laboratory Confirmation

Culturing an ulcer is the ideal way to determine the infecting organism; large central ulcers should be cultured before therapeutic intervention. However, do not delay treatment while waiting for the culture results. With an increase in emerging resistance, many specialists are advocating culture-driven treatment rather than empiric treatment regimens. Consultation with a corneal specialist is recommended if a vision-threatening ulcer is present and/or if there is a risk of perforation or scleral involvement.

Cultures should be obtained any time there is a history of organic trauma, immunocompromised or hospitalized patients, suspicion of an atypical infection, or when an ulcer is unresponsive to seemingly appropriate therapy. Cultures are advised upon initial presentation whenever there is an infiltrate at 25 % depth of the cornea or 50% corneal thinning, regardless of the location.

Table 7. Indications for Cultures.

Mandatory	Recommended	Rarely Helpful
Corneal ulcers, Neonatal conjunctivitis, Hyperacute conjunctivitis, Dacrocystitis	Chronic conjunctivitis, Chronic blepharitis, Hospital infections, Ulcerative conjunctivitis, Epidemic conjunctivitis, Atypical external disease, Follicular conjunctivitis	Dendritic ulcer, Acute blepharitis, Papillary conjunctivitis, Hordeolum/chalazion, Corneal abrasion, Allergic conjunctivitis

When deemed appropriate, various media (blood, or chocolate and Sabouraud’s media if the possibility of fungal infection exists) should be plated directly from a sufficient specimen. (Transport media can be used but will likely reduce the lab's ability to identify the infecting organism.) These media are sufficient to identify most bacteria and fungi. If Mycobacterial sp. is suspected, plating on Lowenstein-Jensen medium should be done and an acid-fast stain should be used.^5,8

Figure 7. Culture showing significant bacterial growth after 72 hours.

Corneal material for culture is best obtained by scraping areas of ulceration with a sterile spatula. Both the edge and central base of the ulcer should be scrapped. This must be done because some organisms (e,g,, Streptococcus pneumoniae and Pseudomonas sp.) are found more readily at the active edge, whereas others (e.g., Moraxella) are found more readily at the base of the ulcer.⁵

Gram stain allows preliminary identification of bacteria causing an infection but should not always serve as a diagnostic guide (especially when a severe infection is present) due to errors in definitively identifying causal organisms by gram positive/negative differentiation.

Stains are also useful if a fungal etiology is suspected. Potassium hydroxide with calcofluor white stain is highly specific and sensitive for fungal detection.⁶

Table 8. Gram Stain Findings.

	Gram Positive Cocci	Gram Negative Cocci	Gram Negative Bacilli (rods)
Examples	Staphylococci, Streptococci, S. pneumoniae, S. aureus, S. pyogenes	Neisseria gonorrhoeae	Pseudomonas aeruginosa, Escherichia, Shigella, Klebsiella, Proteus, Enterobacter, Serratia marcescens, Moraxella

Blood agar (5% defibrinated sheep’s blood in a trypicase-soy agar base) is used for growing the majority of bacteria. To distinguish S. aureus from S. epidermidis, S. aureus produces hemolysis that causes destruction of red blood cells in the agar. This creates clear areas around the colonies. In contrast, S. epidermidis does not produce these clear areas. Neisseria gonorrhoeae does not grow well on blood agar, so chocolate agar (polypeptone agar base enriched with hemoglobin and specific nutrients) is used.⁷

Thioglycolate medium can be used to determine if the bacteria are anaerobic (i.e., those that do not grow readily in the presence of oxygen). A disadvantage of this broth is the inability to monitor and assess contamination, and the fact that it will allow the growth of some aerobes.⁷ Gram negative rods are identified by using different biochemical and fermentation tests such as MacConkey agar.⁷

Sensitivity testing may be helpful, but theses tests are based on systemic values rather than topical levels of antibiotic achieved at the site of infection. Therefore, clinical response should dictate any adjustments in therapy.

Antibiotic susceptibility of bacteria can be tested using two techniques: disc diffusion and antibiotic dilution. Disc diffusion involves inoculating the pathogen on the surface of an agar plate and placing small discs containing various antibiotics on the plate. Bacterial susceptibility is determined by measuring the zone of bacterial growth inhibition surrounding the discs.

Serial dilution testing involves taking antibacterial drugs with various concentrations and inoculating them with the pathogen to determine the minimal inhibitory concentration (MIC) required to prevent bacterial growth.⁷

Table 9. Clinical Features of Bacterial Keratitis.

Gram Positive Organisms	Gram Negative Organisms
Localized Round or oval gray-white infiltrates Distinct borders Minimal surrounding edema	Increased suppuration Adherent exudates to base or edge of ulcer Larger, less defined infiltrate More rapid progression and stromal necrosis/ excavation

Figure 8. Gram positive infection in an aphakic contact lens wearer.

Figure 9. Gram negative infection in a lens patient who wore daily wear lenses overnight.

Specific Bacteria Causing Keratitis

Eighty-seven percent of cases of bacterial keratitis are caused by organisms in the these four groups: Micrococcae (Staphylococcus, Micrococcus), Streptococcus sp., Pseudomonas sp., and Enterobacteriaceae (Citrobacter, Klebsiella, Enterobacter, Serratia, Proteus).¹

Staphylococci: Both S. aureus and S. epidermidis have similar corneal ulcer appearance with a yellow-white round/oval shaped infiltrate having distinct borders, but the tissue surrounding the ulcer margin is often blurred by a stromal infiltrate and edema. S. aureus will cause a more severe microbial ulcer with more complications.⁷

Streptococci: Although not being able to penetrate an intact corneal epithelium, any compromise to that barrier can cause serious corneal infection. S. pneumoniae is often characterized as producing a "serpiginous or creeping" lesion as it spreads toward the central cornea.^5,9 Bacterial exotoxins can cause extensive stromal damage and corneal perforation, as well as a sterile hypopyon.^5,9

Neisseria: Even though not often found in contact lens wearers, N. gonorrhoeae can infect an intact, non-keratinized epithelium. It often occurs as an extension of conjunctivitis that can cause peripheral and central corneal ulcers.¹⁰ A purulent discharge is typically associated with this keratitis. If not treated, it can progress to perforation and panophthalmitis.

Moraxella: Most commonly found in alcoholic or immune-suppressed patients,^11,12 these ulcers are found centrally or inferiorly in areas of exposure. The ulcers are often painless but can cause a large hypopyon if allowed to progress.¹³ Moraxella is sensitive to most antibiotics, but its ulcerative lesions are very slow to heal.

Enterics: Escherichia, Shigella, Klebsiella, Proteus, Enterobacter, and Serratia marcescens have been isolated from corneal ulcers.^10,11 Extended-wear contact lens patients and hard lens wearers with central corneal ulcers have been found to be culture-positive for Serratia marcescens.^13-15

Pseudomonas: Pseudomonas aeruginosa is a common cause of bacterial keratits.⁵ It cannot penetrate an intact corneal epithelium, but infection can follow impairment of the corneal tissue. The corneal infection of P. aeruginosa is the most distinctive of all bacterial corneal infections. It is characterized by a rapid progression of a central or paracentral, broad, shallow ulceration with copious mucopurlent, yellowish-green exudates that are tightly adherent to the conjunctiva and ulcer surface. Perforation is a real threat with P. aeruginosa.¹ The non-ulcerated cornea often has a "ground glass" appearance.

These bacteria often infect contact lens patients from contaminated products such as contact lens solutions or containers or from the hands during contact lens manipulation.⁴ Infection with P. aeruginosa is a serious concern because it can rapidly progress and perforate the cornea within 48 hours. These bacteria produce the enzyme collagenase, so that even if antibiotics kill the bacteria corneal destruction can still continue.³

Treatment of Bacterial Keratitis

When microbial keratitis is suspected, it is important to determine characteristics of the infection such as severity, location, and risk factors (e.g., contact lens wear and previous corneal disease). This assessment should allow the clinician to select an appropriate initial plan including antibiotic use and laboratory studies. Because most bacterial keratitis is caused by one of the four bacterial groups listed above, initial antibiotics should provide coverage for these bacteria.¹

Monotherapy without culturing may be employed if the "1-2-3 rule" is followed:

• 1+ or less anterior chamber reaction
• Infiltrate is less than 2 mm in size,
• The lesion is 3 mm or greater from the visual axis.

All three criteria should be met, otherwise an additional agent to increase broad spectrum coverage should be employed and cultures should be obtained.

Table 10. Guidelines for antibiotics and laboratory studies.

Feature	Suppurative, Severe, Large ulcer	Non-suppurative, Mild-Moderate, Small ulcer
Onset/Progression	Acute/rapid	Sub-acute/slowly progressive
Virulence	Highly likely	Uncertain
Laboratory Studies	Immediate	May delay
Laboratory Materials	Standard	Special
Initial Antibiotics	Based on smears/ broad spectrum antimicrobials	Based on smears and biopsy

Treatment of microbial keratitis should be based on the following principles:

• Broad spectrum initial coverage
• Single agent vs. multiple agent
• Rapid and intensive topical therapy
• Daily evaluation
• Tailoring the antibiotic choice based on culture and clinical response.

These principles aim to eliminate the organism(s) in a timely fashion, reduce inflammation, prevent structural damage, and promote corneal re-epithelialization.¹

The initial choice of antibiotics for the treatment of bacterial keratitis has been altered in recent years by the availability of manufactured topical fluorquinolones. For the past two decades, initial therapy for microbial keratitis involved topical, broad-spectrum, fortified antibiotics such as a first or second generation cephalosporin to cover gram positive cocci and an aminoglycoside to cover gram negative bacteria.¹⁶ However, the initial treatment of choice has now shifted to a more wide-spread use of topical fluoroquinolones. This shift was fueled by studies showing no difference in treatment success between the two therapy approaches.¹⁶

The Ofloxacin Study Group found no statistical difference in cure rate between fortified, dual agent therapy and fluoroquinolone use. In addition, fortified antibiotic treatment was shown to be more toxic to the corneal epithelium (42% vs. 4.3%).¹⁷ Thus, topical fluoroquinolones are now recommended for bacterial keratitis based on their wide-spread availability, long shelf life, broad spectrum coverage, and lack of toxicity.¹⁶ There are some concerns about emerging resistance to topical fluoroquinolone and some reports of increased incidence of corneal perforation (although not verified in clinical trials), but the wide-spread availability of the fluoroquinolones has encouraged empirical treatment and discouraged laboratory investigation.¹⁶

Figure 10. Patient with a significant dermatitis following the use of a fortified aminoglycoside for a corneal ulcer.

Empirical treatment of suspected bacterial keratitis using topical fluoroquinolones in young patients with small ulcers is often sufficient. Results from the Ofloxacin Study showed that young patients with small ulcers do well and are not likely to be culture positive. However, large (over 5 mm) culture-positive ulcers in elderly patients (over 60 years old) had a 5.5 times greater risk of primary treatment failure than did others in the study group.¹⁷ For these large ulcers, a corneal scrape is mandatory to isolate potentially resistant organisms and to identify those that will potentially have a poor response to initial therapy.

Table 11. Distribution of ulcer sizes and their culture results from the Ofloxacin Study.

	25th quartile	Median	75th quartile
Positive culture	2.00 mm2	6.00 mm2	12.00 mm2
Negative culture	0.25 mm2	1.0 mm2	2.25 mm2

There is concern for the emerging resistance of bacteria to earlier generation topical fluoroquinolones, particularly some Streptococcus sp, entercocci, atypical mycobacterium, aeruginosa and non-aeruginosa pseudomonas, and methicillin resistant S. aureus. When resistance has been established, Amikacin may prove to be a reliable topical agent if gram negative organisms have been identified, and Vancomycin can be used when gram positive organisms are suspected.

The following studies describe emerging resistance patterns and trends:

• UCSF Study (1996) Hwuang et al. Using topical ciprofloxacin qid S. aureus resistance ranged from 12%-50% with in-vitro testing.
• Wills Eye Hospital Study (1996) Rodman et al. Staphylococcus aureus resistance decreased from 20% in 1992 to 13% in 1996. Streptococcus sp. resistance increased from 4% in 1992 to 26% in 1996.
• LV Prasad Institute (1999) Kunimoto et al. 30.7% of corneal isolates were not sensitive to ciprofloxacin.
• Bascom Palmer Eye Institute (1999) Chaudry et al. P. aeruginosa resistance rose from 0.44%(1991-1994) to 4.1% (1995-1998).

(Note: In-vitro resistance does not always correlate well with clinical resistance due to the levels of antibiotic achieved at the infection site often being much higher than those obtained in the laboratory setting.)

The use of fourth generation fluoroquinolones has recently gained some acceptance for treatment of bacterial keratitis. These agents provide greater gram positive and equivalent gram negative coverage when compared to earlier generation fluoroquinolones. There appears to be decreased chance for resistance because the newer drugs attack and bind to DNA gyrase (gram negative) and topoisomerase II/IV (gram positive) enzymes essential for bacterial metabolism.

Mechanism of Fluoroquinolone Action

DNA gyrase or topoisomerase-IV are enzymes which enable the bacterial cells' DNA to coil and replicate. They are targeted by fluoroquinolones which prevent replication and transcription, thus preventing bacteria from functioning. Fluoroquinolones bind to the DNA gyrase in gram negative organisms and the topoisomerase-IV enzyme in gram positive organisms.¹³

Fourth-generation antibiotics (gatifloxacin and moxifloxacin) are built on the second-generation molecular structure. An 8-methoxy group is added, the C-7 group is changed, and there is a hydrophobic region to limit bacterial efflux. These new agents are believed to have quorem sensing abilities and perhaps auto-induction potential.

Treatment Regimen For Bacterial Keratitis

Initial treatment with fluoroquinolones should involve rapid and intense therapy. Because of the eye drop administration frequency, patients and families may not be able to comply with the schedule, so hospitalization might be necessary. After an initial loading dose (one drop every 5 minutes for the first 30 minutes), topical antibiotics should be scheduled every 15 min to one hour so the corneal tissue will become rapidly saturated with a high concentration of the drug being applied. This intensive therapy should be continued for 24 to 48 hours day and night.

A concentration exceeding the MIC can be achieved within a few hours and a sustained high concentration for 48 hours is usually sufficient to arrest the infection.² Then, hourly administration for the next 2 to 3 days should sterilize the corneal ulcer. Adding oral doxycycline (doxycycline is a metalloproteinase inhibitor) may prove beneficial, particularly if the ulcer is large and corneal thinning is present. A dosage of 100 mg BID may help to prevent corneal perforation.¹⁸

In conjunction with the antibiotics, a cycloplegic agent should be used to minimize inflammation and increase patient comfort. Additional therapy can include NSAIDs, tissue adhesives, a collagen shield for 12 hours reconstituted in a water-soluble, fortified antibiotics, and subconjunctival injections once or twice daily for 1 to 2 days (rarely if ever needed). Intravenous/oral antibiotics can also be used for impending perforation and scleral suppuration.

Referral to a Corneal Specialist

A corneal specialist referral may be required when there is a lack of clinical response, an exotic organism is involved, or if corneal perforation is a possibility.

A corneal specialist may be especially valuable in the following situations:

• The patient presents with the keratitis after previous surgery
• The infiltrate is over 3.0 mm in diameter
• More than one infiltrate is present
• The infiltrate is near the visual axis
• The patient is immunocompromised

Termination of Therapy

The timing of therapy cessation is often difficult to determine because signs of resolution and healing can be subtle. Criteria for termination can include:

• Blunting of the perimeter of stromal suppuration
• Reduction in density of suppuration
• Reduced cellular infiltrate and surrounding edema
• Reduction in anterior chamber reaction
• Progressive re-epithelialization.

When terminating therapy, it is important to avoid abrupt cessation. After the initial treatment, antibiotic dosing can be reduced to QID.² Healing is often prolonged in large, culture-positive ulcers, especially in the elderly with surface disease.² Topical steroids are often necessary to quiet inflammation and may be needed to promote healing of the epithelial defect.²

Therapeutic response to appropriate levels of antibiotic will vary depending on the organism encountered. Co-infection is always possible and should be considered when seemingly appropriate therapy does not produce a timely response.

Keep in mind that Pseudomonas infections will generally look worse at the 24 hour evaluation and typically take a week or longer to treat because organisms may persist for 14 days or longer. Most gram positive infections respond after 24-48 hours of treatment. Low virulent organisms like S. epidemidis show rapid improvement at 24 hours and S. aureus and S. pneumoniae are relatively unchanged at 24 hours but show rapid improvement at 48 hours. They typically are eliminated at 7-10 days or sooner.

The most treatment-resistant cause of microbial keratitis is acanthamoeba. Infections related to Mycobacterium and filamentous fungi are slowest to respond to appropriate therapy and may persist for weeks.

In rare cases, ulcers do not respond to any available antibiotics or therapy techniques. These cases may require a therapeutic penetrating keratoplasty (PKP). Situations that may require PKP include progressive suppuration, persistent infection, and corneal perforations not manageable by other techniques and procedures. Penetrating keratoplasty may also be needed following resolution of corneal ulceration if vision is significantly affected.

Summary

Bacterial keratitis in contact lens wearers can produce significant corneal morbidity with the potential for devastating vision loss. A timely and accurate diagnosis along with aggressive treatment can help to provide a favorable outcome. Complete assessment of patient history including risk factors and familiarity of clinical presentation will help in making an accurate diagnosis. The use of newer generation fluoroquinolones has helped in the treatment of bacterial keratitis, but dual agent therapy along laboratory confirmation remains important in the management of significant corneal infections presented by patients who wear contact lenses.

References

1. Kaufman HE, Barron BA, McDonald MB. The Cornea. Churchill Livingstone, New York 1988; 217-247.

2. Morlet N, Daniell M. Empirical fluoroquinolone therapy is sufficient initial treatment. Br J Ophthalmology 2003; 87: 1169-1172.

3. Swoka J, Gurwood A, Habat A. The Handbook of Ocular Disease Management. Suppl. To Review of Optometry. March 1997 18A-19A.

4. Dewart MR, Elliot LJ. Management of Contact Lens Associated or Lens Induced Pathology. Clinical Manual of Contact Lenses. Eds. Bennett ES, Henry VA, Lippincott Williams and Wilkins. Philadelphia 2000. 582-610.

5. Abbot RL, Abrams MA. Bacterial Corneal Ulcers. In W Tasman, EA Jaeger (eds), Duane’s Clinical Ophthalmology (2nd ed). Philadelphia: Lippincot, 1989;Vol 4 (18):1-34.

6. Sharma s, Kunimoto DY, Gopinatha U, Athmanatha S, Garg P, Rao GN. Evaluation of corneal scraping smear examination methods in the diagnosis of bacterial and fungal keratitis: a survey of eight years of laboratory experience. Cornea 2002 21;7: 643-7.

7. Stern GA. Contact Lens Associated Bacterial Keratitis: Past, Present, and Future, CLAO J 1998, 24:52-56.

8. Broadway DC, Kerr-Muir MG, Eyyn SJ, Pambakian H. Myocbacterium chelonei Keratitis. A case report and review of previously reported cases. Eye 1994;8: 134-142.

9. Kaufman HE, Barron BA, McDonald MB, Kaufman SC. The Companion Handbook to the Cornea 2nd Edition. Butterworth-Heinemann, Boston 2000;91-145.

10. Liesegang TJ, Forster RK. Spectrum of Microbial Keratits in South Florida. Am J Ophthalmology 1980;90:38-47.

11. Yolton DP. Ocular Microbiology and Cytology. In Roberts DH, Terry JE. Ocular Disease 2nd Ed 1996; 373-399.

12. Baum J, Fedukowicz HB, Jordan A. A Survey of Moraxella corneal ulcers in derelict population. Am J Ophthalmology 1980;90:476-80.

13. Feducowicz HB, Horwich H. The gram-negative diplobacillus in hypopyon keratitis. Arch Ophthalmol 1953;49:202-11.

14. Templeton WC III, Eiferman RA, Synder JW, et al. Serratia keratitis transmitted by contaminated eyedroppers. Am J Ophthalmology 1982;93:723-6.

15. Lass JH, Haaf J, Foster CS, Belcher C. Visual outcomes in eight cases of Serratia marcescens keratitis. Am J Ophthalmology 1982;92:384-90.

16. Daniell M, Mills R, Morlet N. Veiw 1 Microbial keratitis: what’s the preferred initial therapy?. Br J Ophthalmology 2003; 87: 1167-1169.

17. Ofloxacin Study Group. Ofloxacin monotherapy for the primary treatment of suspected microbial keratitis. Ophthalmology 1997; 104: 1902-9.

18. Ralph RA. Tetracyclines and the treatment of corneal stromal ulceration. Cornea 2000;19: 274-7.

Contact the Author:

• Joseph P. Shovlin, OD, FAAO
• Northeastern Eye Institute
• 200 Mifflin Ave.
• Scranton PA 18503
• JShovlin@aol.com

Pacific University College of Optometry provides On-Line CE as a service to optometrists. The college does not endorse or recommend any products, equipment, or services that might be discussed in the courses. Courses are prepared by individuals believed to be experts in their areas of specialization who are compensated for their efforts. The College relies on their expertise to produce accurate and timely courses. Questions or concerns about courses should be directed to the individual authors and/or the Continuing Education Department at the College of Optometry at kundart@pacificu.edu .

© Copyright 2006, Pacific University College of Optometry