Rapid onset of pain is accompanied by conjunctival injection, photophobia, and decreased vision in patients with bacterial corneal ulcers. The rate of progression of these symptoms depends on the virulence of the infecting organism. Bacterial corneal ulcers are typically a single infiltrate and show a sharp epithelial demarcation with underlying dense, suppurative stromal inflammation that has indistinct edges and is surrounded by edema. P aeruginosa typically causes stromal necrosis with a shaggy surface and adherent mucopurulent exudate (Fig 10-15). An endothelial inflammatory plaque, marked anterior chamber reaction, and hypopyon frequently occur.
Figure 10-15 Suppurative ulcerative keratitis caused by P aeruginosa.
Patients with infections caused by slow-growing, fastidious organisms such as mycobacteria or anaerobes may have a nonsuppurative infiltrate and intact epithelium. Infectious crystalline keratopathy, for example, presents as densely packed, white, branching aggregates of organisms in the virtual absence of a host inflammatory response, shielded by the bacterial biofilm coating. Risk factors include corticosteroid use, contact lens wear, and previous corneal surgery. Infectious crystalline keratopathy has been reported with a number of bacterial and fungal species, most commonly α-hemolytic Streptococcus species (Fig 10-16).
The prevalence of a particular causative organism depends on the geographic location and risk factors for the infection. Causative organisms in bacterial keratitis are listed in Table 10-5.
Studies indicate that for bacterial keratitis, clinical appearance of the infection is an unreliable guide in determining the causative pathogen. The successful use of topical fluoroquinolones in the 1990s led to a reduction in the number of cultures performed for cases of presumed infectious keratitis. The American Academy of Ophthalmology practice guidelines continue to recommend that initial cultures be obtained for infiltrates extending to the middle of the cornea, into deep stroma, or across a large area (>2 mm), as well as for patients whose history or clinical features suggest fungal, amebic, mycobacterial, or drug-resistant organisms as the causative agents. In addition to corneal culture, it may be helpful to culture contact lenses, contact lens cases and solutions, and any other potential sources of contamination, such as inflamed eyelids. Some correlation has been shown between cultures of such sources and corneal scrapings.
Figure 10-16 Infectious crystalline keratopathy in a corneal graft caused by α-hemolytic Streptococcus species.
Table 10-5 Causes of Bacterial Keratitis
The yield for corneal cultures and smears is significantly higher before the initiation of antibiotic treatment, but cases unresponsive to such therapy should still be cultured, with some suggesting discontinuation of antibiotics for 12–24 hours to encourage yield. However, a positive smear result at any point does not obviate the need for broad-spectrum coverage, although it may cause coverage to be weighted toward a different class of microorganism and/or provide guidance for later treatment in the absence of a positive culture. (See Chapter 9 in this volume and BCSC Section 4, Ophthalmic Pathology and Intraocular Tumors, for discussion of specimen collection, culturing, staining, and interpretation.)
In any keratitis, the primary goal of therapy is preservation of sight and corneal clarity. Bacterial pathogens can produce irreversible corneal scarring over a period of hours because of their rapid growth, keratolytic enzymes, and stimulation of destructive host immune responses. Therefore, therapy must be initiated before definitive diagnosis is obtained in order to rapidly reduce the bacterial load and minimize later visual disability.
Initial therapy consists of empiric, broad-spectrum topical antibiotics. In routine corneal ulcers, monotherapy with topical fluoroquinolones provides outcomes equivalent to those of combination therapy, because of the excellent penetration achieved with commercially available concentrations of fluoroquinolones. These antibiotics should initially be given every 30–60 minutes and then tapered in frequency according to the clinical response. In severe cases, administration of antibiotics every 5 minutes for 30 minutes as a loading dose can more rapidly achieve therapeutic concentrations in the corneal stroma. Second-generation fluoroquinolones (ciprofloxacin, ofloxacin) continue to have excellent Pseudomonas coverage but lack useful gram-positive activity. Third- and fourth-generation fluoroquinolones (eg, moxifloxacin, gatifloxacin, levofloxacin, and besifloxacin) have improved gram-positive and atypical mycobacterial coverage but limited activity against MRSA.
Alternatively, topical combination therapy with an agent active against gram-positive bacteria and another agent active against gram-negative bacteria can be used as initial therapy (Table 10-6). Although “fortified” antibiotics (compounded at increased concentrations compared with their commercial formulations in order to achieve therapeutic levels in the corneal stroma) are more difficult to obtain and may have a greater toxic effect on the ocular surface, the clinician should consider using them, especially in combination with vancomycin for gram-positive coverage when MRSA is suspected, with large or vision-threatening ulcers, or with prior antibiotic failure. Effectively treated, most infectious keratitis is culture negative after 48–72 hours. Once the offending microbe is identified or the clinical response shows improvement, appropriate monotherapy may be considered (see Table 10-6) to maintain coverage and reduce toxicity. However, laboratory sensitivities are based on antibiotic tissue levels achievable by systemic administration, and the levels achieved by topical administration are much higher. Often, a bacterial keratitis will respond in vivo even when in vitro data suggest resistance. Any changes in medical therapy should therefore be based primarily on clinical response. Several clinical parameters are useful for monitoring clinical response to antibiotic therapy:
Table 10-6 Initial Therapy for Bacterial Keratitis
blunting of the perimeter of the stromal infiltrate
decreased density of the stromal infiltrate
reduction of stromal edema and endothelial inflammatory plaque
reduction in anterior chamber inflammation
cessation of corneal thinning
Systemic antibiotics—especially the fluoroquinolones, which have excellent ocular penetration—and intensive topical antibiotics are indicated in cases with suspected scleral and/or intraocular extension of infection.
The role of corticosteroid therapy for bacterial keratitis remains controversial. Tissue destruction results from a combination of the direct effects of the bacteria and an exuberant host inflammatory response consisting of polymorphonuclear leukocytes and proteolytic enzymes, which predominate even after corneal sterilization. Corticosteroids are effective at modifying this response, but they also inhibit the host response to infection. The literature strongly suggests that corticosteroid therapy administered prior to appropriate antibiotic therapy worsens prognosis. The literature is inconclusive, though, about steroid therapy used concomitantly with antibiotic therapy or after it is initiated, as demonstrated in a randomized clinical trial in which topical corticosteroids were given 48 hours after initiation of topical antibiotics for bacterial keratitis. At 3 months, no effect on final visual outcome or complication rate was seen, but a trend toward improved outcomes was noted in those patients with the worst initial vision who received corticosteroids and for the corticosteroid group at 1-year follow-up. Notably in this study, Nocardia keratitis, which is uncommon in the United States, fared worse with corticosteroid treatment.
The indiscriminate or universal use of corticosteroids is, therefore, unsupported but does not appear to increase the general risk of poor outcomes or complications in treated bacterial keratitis. In fact, certain patients may benefit from the addition of corticosteroids to antibiotic therapy. Future study of the appropriate timing and dosage may further refine the indications for corticosteroid use. As there is still significant risk associated with corticosteroid use in patients with bacterial or other forms of infectious keratitis not appropriately treated, following are recommended criteria for instituting corticosteroid therapy for bacterial keratitis:
Corticosteroids should not be used in the absence of appropriate antibiotic therapy.
The patient must be able to return for frequent follow-up examinations and demonstrate adherence to appropriate antibiotic therapy.
No other associated virulent or difficult-to-eradicate organism is found or suspected.
Corticosteroid drops may be started in moderate dosages (prednisolone acetate or phosphate 1% every 6 hours), and the patient should be monitored at 24 and 48 hours after initiation of therapy. If the patient shows no adverse effects, the frequency of administration may be adjusted based on clinical response. Collagen crosslinking is increasingly used as an adjunctive therapy for bacterial keratitis, with anecdotal success; as this technology becomes more available in the United States, its precise role and application are evolving.
Penetrating keratoplasty (PK) for treatment of bacterial keratitis is indicated if the disease progresses despite therapy, descemetocele formation or perforation occurs, or the keratitis is unresponsive to antimicrobial therapy. The involved area should be identified preoperatively and an attempt made to circumscribe all areas of infection. Peripheral iridectomies are indicated, because seclusion of the pupil may develop from inflammatory pupillary membranes. Interrupted sutures are recommended. The patient should be treated with appropriate antibiotics, cycloplegics, and intense topical corticosteroids postoperatively. See Chapter 15 in this volume for a more detailed discussion of PK and BCSC Section 2, Fundamentals and Principles of Ophthalmology, for an in-depth discussion of ocular pharmacology.
American Academy of Ophthalmology Cornea/External Disease Panel. Preferred Practice Pattern Guidelines. Bacterial Keratitis. San Francisco: American Academy of Ophthalmology; 2013. Available at www.aao.org/ppp.
Price MO, Tenkman LR, Schrier A, Fairchild KM, Trokel SL, Price FW Jr. Photoactivated riboflavin treatment of infectious keratitis using collagen cross-linking technology. J Refract Surg. 2012;28(10):706–713.
Schein OD, Glynn RJ, Poggio EC, Seddon JM, Kenyon KR. The relative risk of ulcerative keratitis among users of daily-wear and extended-wear soft contact lenses. A case-control study. Microbial Keratitis Study Group. N Engl J Med. 1989;321(12):773–778.
Srinivasan M, Mascarenhas J, Rajaraman R; Steroids for Corneal Ulcers Trial Group. Corticosteroids for bacterial keratitis: the Steroids for Corneal Ulcers Trial (SCUT). Arch Ophthalmol. 2012;130(2):143–150.
Srinivasan M, Mascarenhas J, Rajaraman R; Steroids for Corneal Ulcers Trial Group. The steroids for corneal ulcers trial (SCUT): secondary 12-month clinical outcomes of a randomized controlled trial. Am J Ophthalmol. 2014;157(2):327–333.